‘
~
-’
.
NE~=~A~~~~FNo~~~~‘cE-:
-
SAFETY
BOARD
AIRCRAFT ACCIDENT
SUPERSEDES
STATES. GOVERNMENT
TECHNICAL
22161
..
5
’s
‘%%&%~p
’CR
”.
-
__
.
TECHNICAL REPORT DOCUMENTATION
PAGE
.
Catalog
.
e ano
Aircraft Accident Report-Air
Canada Flight 797, McDonnell Douglas DC-9-32, C-FTLU,
Greater Cincinnati International Airport, Covington,
June 2, 1983
(SuDersedes NTSB/AAR-84/09)
1986
6.Performing Organization
8.Performing Organization
National Transportation Safety Board
Bureau of Accident Investigation
Washington, D.C. 20594
Aircraft Accident Report
June 2, 1983
TRANSPORTATION SAFETY
On June 2, 1983, Air Canada Flight 797, a McDonnell Douglas DC-9-32, of
Canadian Registry C-FTLU, was a regularly scheduled international passenger flight from
Dallas, Texas, to Montreal, Quebec, Canada, with an en route stop at Toronto, Ontario,
Canada. The flight left Dallas with 5 crewmembers and 41 passengers on board.
About 1903, eastern daylight time, while en route at flight level 330 (about
33,000 feet
the cabin crew discovered a fire in the aft lavatory. After contacting air
traffic control
and declaring an emergency, the crew made an emergency descent, and
ATC vectored Flight 797 to the Greater Cincinnati International Airport, Covington,
Kentucky.
At
eastern daylight time, Flight 797 landed on runway 27L at the
Greater Cincinnati International Airport.
As the pilot stopped the airplane, the airport fire
department, which had been alerted by the tower of the fire on board the incoming plane, was
in place and began firefighting operations.
Also, as soon as the airplane stopped, the flight
attendants and passengers opened the left and right forward doors, the left forward
exit, and the forward and aft right exits.
About 60 to 90 seconds after the exits
were opened, a flash fire enveloped the airplane interior.
While 18 passengers and 3 flight
attendants exited through the forward doors and slides and the three
exits to to
’d
_
--
Abstract
On August 8, 1984, the National Transportation Safety Board adopted the report
and probable cause of the accident.
On December 20, 1984, the Air Line Pilots
Association submitted a petition for reconsideration of the contributing factors statement
of the probable cause that was adopted in the original report.
As a result of the Air Line
Pilots Association’s petition,
the accident report and the probable cause have been
revised.
The National Transportation Safety Board determines that the probable causes of
the accident were a fire of undetermined origin, an underestimate of fire severity, and
misleading fire progress information provided to the captain.
The time taken to evaluate the nature of the fire and to decide to initiate an
emergency descent contributed to the severity of the accident.
;
.
CONTENTS
1
SYNOPSIS . . . . . . . . . . . . . . . . . . . , . . . . . .
,
1.
FACTUAL INFORMATION.
..................
2
1.1
History of the Flight .....................
2
1.2
Injuries to Persons ......................
8
1.3
Damage to Airplane ......................
9
1.4
Other Damage ........................
9
1.5
Personnel .....................
9
1.6
Airplane Information .....................
9
1.6.1
Flight and Cabin Maintenance Logbook Writeups
... .
..... .
10
1.6.2
Passenger Cabin Modification.
................ .
11
1.7
Meteorological Information. ..................
12
Aids to Navigation ......................
12
1.9
Communications
.......................
12
1.10
Aerodrome Information .....................
12
1.11
Flight Recorders
.......................
13
1.12
Wreckage and Impact Information
............... .
14
1.12.1
External Fuselage
.......................
14
1.12.2
Interior Fuselage Forward of the Aft Lavatory
......... .
16
1.12.3
Aft Lavatory Area ......................
16
1.12.4
Cargo and Aft Accessory Compartments
............ .
25
1.12.5
Cockpit Controls and
............... .
27
1.13
Medical and Pathological Information
............. .
27
1.14
Fire Response
........................
28
1.15
Survival Aspects
.......................
30
1.16
Tests and Research ......................
35
1.16.1
Federal Bureau of Investigation (FBI) Laboratory Tests.
..... .
35
1.16.2
Electrical System Components ................ .
35
1.16.3
Flush Motor and Lavatory Components. ............ .
36
1.16.4
Flush Motor Seizure Test. ...................
38
1.16.5
Fire and Heat Tests ......................
38
1.16.6
Airplane Cabin Fire Research. ................ .
42
1.17
Other Information. ......................
43
1.17.1
Air Canada Operational Procedures
.............. .
43
l.17.2
Aft Lavatory. .......................
46
1.17.3
Smoke Detectors
.......................
47
1.17.4
Examination of Other DC-9 Airplanes
............. .
47
,
1.17.5
Air Traffic Control Procedures ................ .
47
1.17.6
Air. Traffic Control Radar Data ................ .
50
1.17.7
DC-g-32 Descent Performance ..................
51
2.
ANALYSIS
......................... .
53
2.1
General
............................
53
2.2
Fire
.............................
53
2.3
Operational and Survival Factors
............... .
59
2.4
Firefighting
..........................
68
2.5
In-flight Fire Prevention/Detection
................
6 8
3.
CONCLUSIONS.
.......................
69
3.1
Findings
.
. .
...............................................
69
3.2
Probable Cause ......
7 1
111
cause.
................
.-.
..
..-.
.....
i
.
9 9
4.
.
. . . . . . , . . . . . . . . . . . . .
71
5.
APPENDIXES
........................
79
Appendix A-Investigation and Hearing.
............ .
79
Appendix B-Personnel Information
.............. .
80
Appendix C-Airplane Information.
.............. .
82
Appendix D-Cockpit Voice Recorder Transcript
........ .
83
Appendix E-14 CFR 25 Flame Resistance Criteria
....... .
97
Appendix F-Air Canada DC-9 Emergency Equipment
...... .
98
Appendix G--Petition for Reconsideration of Probable
Appendix
Response to Petition for Reconsideration of
Probable Cause
.......................
109
~
NATIONAL +RANSPORTATION
SAFETY HOARD
.
WASHINGTON, D.C. 20594
AIRCRAFT ACCIDENT REPORT
.
Adopted: January
AIR CANADA FLIGHT 797
M
C
DONNELL DOUGLAS DC-9-32, C-FTLU
GREATER CINCINNATI INTERNATIONAL AIRPORT
COVINGTON, KENTUCKY
JUNE
SYNOPSIS
On June 2, 1983, Air Canada Flight 797, a McDonnell Douglas DC-9-32, of
Canadian Registry C-FTLU, was a regularly scheduled international passenger flight from
Dallas, Texas, to Montreal, Quebec, Canada, with an en route stop at Toronto, Ontario,
Canada. The flight left Dallas with 5 crewmembers and 41 passengers on board.
About 1903, eastern daylight time, while en route at flight level 330 (about
33,000 feet
the cabin crew discovered smoke in the left aft lavatory. After
attempting to extinguish the hidden fire and then contacting air traffic control
and
declaring an emergency, the crew made an emergency descent and ATC vectored Flight
797 to the Greater Cincinnati International Airport, Covington, Kentucky.
.
At eastern daylight time, Flight 797 landed on runway 27L at the
Greater Cincinnati International Airport. As the pilot stopped the airplane, the airport
fire department, which had been alerted by the tower to the fire on board the incoming
plane, was -in place and began firefighting operations.
Also, as soon as the airplane
stopped, the flight attendants and passengers opened the left and right forward doors, the
left forward
exit, and the right forward and aft exits. About 60 to 90
seconds after the exits were opened, a flash fire engulfed the airplane interior. While 18
passengers and 3 flight attendants exited through the forward doors and slides and the
three open
exits to evacuate the airplane, the captain and first officer exited
through their respective cockpit sliding windows.
However, 23 passengers were not able
to get out of the plane and died in the fire. The airplane was destroyed.
The National Transportation Safety Board determines that the probable causes
of the accident were a fire of undetermined origin, an underestimate of fire severity, and
misleading fire progress information provided to the captain.
The time taken to evaluate the nature of the fire and to decide to initiate an
emergency descent contributed to the severity of the accident.
”
to
”
Excuse
’d
”
have
”
it’s
did
”
”
I
”
He then proceeded saturate the washroom with
CO
by spraying the paneling and
.
the seam from which smoke was seeping and spraying t
a
e door of the trash bin. He then
closed the lavatory door.
At the No. 2 flight attendant reached the cockpit and told the
captain, me, there’s a fire in the washroom in the back, they’re just. . back
to go to put it out.”
Upon being notified of the fire, the captain ordered the first officer
to inspect the lavatory.
The captain then donned’ his oxygen mask and selected the
oxygen position on his regulator.
The first officer left the cockpit but did
not take either smoke goggles or a portable oxygen bottle with him. (The airplane was not
equipped with nor was it required to be equipped with self-contained breathing equipment
a could not get to the aft lavatory
because the smoke, which had migrated over the. last., three to four rows of seats, was too
thick. The flight attendant in charge told the first officer what he had seen when he
opened the lavatory door, that he had discharged the
extinguisher into the lavatory,
and that he had not been able to see the source of the smoke before closing the door.
He
told the first officer, however, that he did not believe the fire was in the lavatory’s trash
container. The first officer told the flight attendant in charge that he was going forward
to get smoke goggles.
At
the first officer returned to the cockpit and told the captain that
the smoke had prevented him from entering the aft lavatory and that he thought
better go He did not tell the captain that the flight attendant in charge had told
him that the fire was not in the trash bin. However, at
before the captain could
respond, the flight attendant in charge came to the cockpit and told the captain that the
passengers had been moved forward and that the captain didn’t
to worry, I think its
gonna be easing
The first officer looked back into the cabin and said that it was
almost clear in the back. At
he told the captain, starting to clear now,” and
that he would go aft again if the captain wanted him to do so. According to the captain,
the first officer’s smoke goggles were stored in a bin on the right side of the cockpit and
were not easily accessible to the first officer while he was not in his seat. Since the first
officer needed the goggles and since there was a hurry, the captain gave him his goggles
and, at
directed him to go aft.
The first, officer also testified that the captain
and he
not discuss the type of fire at during the time he was in the cockpit before
he went to the lavatory the second time.
At
while the first officer was out of the cockpit, the flight attendant
in charge told the captain again that the smoke was clearing. The captain testified that
he believed the fire was in the lavatory trash bin and that he did not decide to descend at
this time because,
expected it (the fire) to be put
In the meanwhile, the first officer proceeded to the aft lavatory and put on
the smoke goggles. He testified that he had intended to open the door to see what the
situation was inside, but when he discovered that the lavatory door felt hot to the touch,
he decided not to open it and instructed the cabin crew to leave it closed. At that time,
he noticed a flight attendant signaling him to hurry back to the cockpit. The first officer
returned to the cockpit and got into his seat, and at
he told the captain, don’t
like what’s happening, I think we better go down,
okay
The captain testified that, from
the first officer’s voice inflection, he knew that the first officer believed the fire was out
of control and that he had to descend immediately.
About
while the first officer was aft to inspect the aft lavatory, the
airplane had experienced a’ series of ‘malfunctions:
According to the captain,
the master caution light illuminated, indicating that the airplane’s left a.c. and
’
code
”
an
”
electrical systems had lost power. At the captain called Indianapolis Center and
requested the Center to standby because the flight had an “electrical problem.” About 30
to 45 seconds later, the Louisville high radar sector controller working Flight 797 lost the
flight’s radar beacon target. The controller then directed the computer to track all
primary targets. Flight 797’s position was
depicted on the scope by a plus sign and
associated data block.
About
after the first-officer had returned to the cockpit, the master
warning light illuminated and the annunciator lights indicated that the emergency a.c.
and d.c. electrical buses had lost power.
The captain’s and first officer’s attitude direc-
tional indicators tumbled.
The captain ordered the first officer to activate the
emergency power switch, thereby directing battery power to the emergency a.c. and d.c.
buses. The attitude directional indicators’ gyros began erecting, however, because of the
loss of a.c. power, the stabilizer trim wasinoperative and remained so during the rest of
the flight.
The descent
At the Air Traffic Control Facilities.-At
Flight 797 called the radar
high sector controller at Indianapolis Center and said,
“Mayday, Mayday, Mayday.” The
Louisville radar high sector controller acknowledged the call, and at
the flight
told the controller that it had a fire and was going down. The controller told the flight
that it was 25 nautical miles
from Cincinnati and asked “can you possibly make
Cincinnati.”
The flight answered that it could make Cincinnati and then requested
clearance; it was then cleared to descend to 5,000 feet. At
Flight 797 reported
that it was leaving FL 330. The flight then told the controller that it needed to be
vectored toward Cincinnati, that it was declaring an emergency, and that it had changed
its transponder code to 7700 -- the emergency code.
However, the transponder was
inoperative due to the power loss, and the emergency code was never portrayed on the
Center’s radarscopes.
At
the Louisville radar high sector controller directed the
flight to turn to
and told it that the Greater Cincinnati Airport (Cincinnati Airport)
was at “twelve o’clock at twenty miles.”
The controller said that it was obvious to him
.
that Flight 797 had to descend *‘immediately
therefore, he issued the clearance and
stated that he was going to coordinate the descent with the other sectors at the center
later. He further stated that the
was intended to place the flight on course
toward Cincinnati Airport.
He heard Flight 797 report leaving FL 330; however, because
of the inoperative transponder, mode C altitude information was no longer being received
and there was no indication on his radarscope that the flight was descending.
At
Indianapolis Center’s Lexington low altitude D (LEX-D) controller
called the approach controller at the Cincinnati Airport’s Terminal Radar Control
facility to alert him of an impending handoff in his southwest sector. Six
seconds later, the LEX-D controller told the approach controller at Cincinnati he had a
for
and at the LEX-D controller then told the Cincinnati approach
controller that he had
emergency for
Air Canada
approach controller replied, “Zero six six two, thirty-five thousand.” Zero six six two was
the code assigned to Continental 383, a westbound flight
FL 350. At the
LEX-D controller answered,
‘*Yeah, thirty-three right now, he’s twenty-five southwest.”
The approach controller replied, “Radar contact, okay.”
However, at
when the
approach controller accepted the handoff of Flight 797, he had mistaken the radar beacon
The international radiotelephonic distress signal. When repeated three times, it
imminent and-grave danger-and that immediate assistance isrequested.
’
’.
’t
say
’s
*fully
’
’s
’s
’
‘I
-
’
-5-
target of Continental Flight 383 as that of Flight 797. Shortly after he had accepted the
handoff, the approach controller had notified the Cincinnati Airport tower local controller
that he intended to land an Air Canada jet with an on board fire on runway 36. The
tower’s local controller alerted the airport fire station, and crash-fire-rescue
vehicles were dispatched and positioned for an emergency landing. The firefighters had
also been advised that the airplane had electrical problems, that smoke was coming from
the aft lavatory, and that there was smoke or fire in the rear of the airplane.
At
almost coincident with the end of his message to the local controller, the LEX-D
controller informed the approach controller of Flight
assigned heading.
Although the approach controller repeated the heading, he stated that he could not
hearing this
At Flight 797 the Cincinnati, approach controller, declared
an emergency, and said that it was descending.
The approach controller acknowledged
and told the flight to plan for a runway 36 instrument landing system
approach and
requested the flight to turn right to
He then realized the target he was observing
was not responding and attempted unsuccessfully to assign a discrete transponder code to
it in order to track it better. Thereafter, at
Flight 797 reported that it had a
fire in its aft lavatory and that the cabin was filling with smoke. The controller asked the
flight to
the type airplane, number of people on board, and amount of fuel (on
board).” The first officer answered that he would supply this data later because
have time now.**
At the approach controller called the D controller on
the
to request assistance. Almost simultaneous with the call, he also observed an
eastbound primary target and began to monitor it. At
the flight requested the
cloud ceiling at the airport and the controller responded that the ceiling was
thousand five hundred scattered, measure(d) eight thousand feet overcast, visibility one
two miles with light rain
The controller then decided that the eastbound target was
Flight 797, and at
he requested the flight to altitude.** The approach
controller said that, by
he knew that he was observing Flight primary
target, but
that it was not identified.
He also knew, based on the target’s position
-- about 3 nmi east of runway
extended centerline and about 8 nmi south of its
threshold -- and its reported altitude of 8,000 feet, that it was too high and too fast to
land on runway 36. He decided to use runway 27L for landing, and used the primary target
to monitor the flight and vector it toward the airport.
At
after Flight 797 was unable to tell him its heading because its
heading instruments were inoperative, the approach controller asked the flight to turn
left. The controller said that this was an identification turn and that it was also designed
to place the airplane closer to the airport. At
after observing the target in a left
turn, the approach controller said that Flight primary target was now **fully
identified.”
He then told the flight that this was a
*ho gyro*
radar approach for runway
two seven left
. .
and cleared it to descend to 3,500 feet. He then told the flight that it
was 12 nmi southeast of the Cincinnati Airport, cleared it to land on runway
and
informed it that the surface wind was
at 4 knots. He informed the tower of the
change of landing runways and the tower directed
the fire department to position its
vehicles along runway 27L. (See figure 1.)
No gyro approach/vector
A radar approach/vector provided in case of a
malfunctioning gyro compass or directional gyro.
Instead of providing the pilot with
headings to be flown, the controller
the radar track and issues control
instructions right/left** or ‘stop as appropriate.
--
’
TOWER DRIVE
FAA CONTROL TOWER
CRASH/FIRE/RESCUE
STATION
967
Figure Positidns of Units.
.
’t
’
’then
’s
‘I
-7-
At Flight 797 reported that it was level at 2,500 feet, and at
that it was [visual flight rules] now. . .
The approach controller vectored the
flight to runway and at told it that it was 12 nmi from the airport. The
flight descended to 2,000 feet, and the controller continued to supply range calls.
At the controller told the flight that the crash-fire-rescue vehicles
were standing by and again asked the flight to provide the number of persons and the
amount of fuel on board. Flight 797 answered, **We
have time, its getting worse
here.**
At the runway and approach lights were turned up to full intensity.
At
Flight 797 reported the airport in sight; the approach controller cleared it to
land, and told it that the surface wind was at 4 knots. At the approach
controller told the flight that it was.3 from asked the tower local
controller if she had the airplane in sight. The local controller said that she did. After
telling Flight 797 that it was 2 nmi from the airport, the approach controller asked the
local controller to tell him when Flight 797 had landed. At
the local controller
told the approach controller,
landed.**
On the
As
the airplane descended, the smoke front continuously
moved forward filling the passenger cabin and entering the cockpit. The first officer had
left the captain’s smoke goggles in the aft end of the airplane. Therefore, after he
returned to the cockpit, he gave one of the two pairs of smoke goggles stowed on the right
cockpit wall to the captain. (See appendix
F.) .Although
there was another pair available,
the first officer chose not to use them.
The captain donned the smoke goggles and wore
them during
the descent and landing.
The captain said
that during the descent, he wore
his oxygen mask and the oxygen regulator was set to the loo-percent position; therefore,
he had no trouble breathing. However, during the latter stages of the approach and
landing,
he had difficulty seeing the instruments because of the smoke in the cockpit and
had to lean forward to do so.
He testified that his perspiration was causing his smoke
goggles to steam up and he had to pull them away from his face from time to time to
clear them. The first officer also wore his oxygen mask during the descent and set his
oxygen system regulator to the loo-percent position and encountered no trouble in
breathing.
The captain began the emergency descent almost simultaneously with the
call. The throttles were retarded to idle and the speed brakes were extended.
However, when the speed brakes were deployed, the spoiler/speed brake handle was moved
inadvertently to the full aft position and the spoiler panels were deployed to the full-up or
ground position. The captain testified that this had no effect on the airspeed during the
descent but it increased the descent rate. The descent was flown at 310 knots indicated
airspeed
and since the needle on the instantaneous vertical speed indicator
was pegged, the rate of descent exceeded 6,000 feet per minute
According to the flightcrew, Flight 797 was operating in visual meteorological
conditions before the emergency descent.
The captain said that the airplane was almost
totally in clouds from about FL 250
to about 3,000 feet; however, it did not encounter
either turbulence or icing.
At
at 3,000 feet, the airplane was in and out of the
cloud bases so he descended to 2,000 feet to obtain VFR flight conditions. According to
the flightcrew, except for the cloud conditions at 3,000 feet, the
descent and
landing were not affected
by weather.
1
-
-9-
Figure 2.-Airplane after fire burned through top of fuselage.
1.3
Damage to
Airplane
The airplane was destroyed by
1.4
Other Damage
None.
1.5
Information
The flightcrew were certificated and the flight attendants were qualified in
accordance with current Canadian regulations.
(See appendix B.) Air traffic control
(ATC) controllers were qualified in accordance with current United States regulations.
Information
1.6
The airplane, a McDonnell Douglas DC-9-32, Canadian registry C-FTLU, was
owned and operated by Air Canada, a Canadian Crown corporation.
The airplane was
manufactured on April 7, 1968, and had been operated by Air Canada since delivery to the
company at that time. (See appendix C.)
”
’
-
-
On September 17, 1979, the airplane experienced an in-flight failure of its aft
pressure bulkhead shortly after takeoff from Logan International Airport, Boston,
Massachusetts. The separation and ensuing occurred shortly after the
airplane had leveled off at FL 250. At the time of the Logan accident, the airplane had
flown about 28,425 hours and had completed 26,816 landings. The damage to the aft part
of the airplane was extensive.
There was disruption of some engine and flight
components. Except for severed flight data recorder connections, no damage was found
on any electrical components,
wires, and cables examined during the investigation.
However, in effecting repairs, numerous wire bundles were cut in order to examine the
airplane and to facilitate the removal of damaged structure and reinstallation of
replacement structure. Repairs to the airplane were made by McDonnell Douglas and
inspected by Air Canada under its authority as a Canadian Ministry of Transport (MOT)
approved company.
The aft pressure bulkhead and aft accessory compartment were
rebuilt at Logan Airport between September 18, 1979, and November 20, 1979. The
installation of the aft lavatories and interior furnishings was made by Air Canada at their
Dorval Base in Montreal.
Air Canada and McDonnell Douglas each wrote engineering
reports on repairs to the airplane. An FAA Form 337, which was part of the Air Canada
report
listed 29 individual repair items.
Item 3 of this list stated, “Spliced electrical
wires through aft pressure bulkhead per service sketch
The sketch designated
where the splices were to be made and the manner in which they were to be made.
addition, the Air Canada report stated that the contractual agreement required the
.
repairs be carried out to restore the aircraft to condition substantially conforming to
specification for the airplane as originally delivered.**
The airplane was returned to
service December 1, 1979.
During the investigation of the Cincinnati accident, all of the wire splices
made during the repairs at Logan which were found and not destroyed were examined. No
evidence of arcing or short circuiting was found.
Flight and Cabin Maintenance Logbook Writeups
Between June 1, 1982, and June 2, 1983, 76 writeups were entered in the
airplane flight log concerning the two engine-driven generators and the auxiliary power
unit
generator.
One writeup concerned the right engine-driven generator; 6
writeups concerned the left engine-driven generator; 34 concerned the APU generator;
and 35 concerned
relay lockout malfunctions.
Of the 35 writeups relating to the
relay lockout, only 1 on December 30, 1982 described an accompanying
generator malfunction. The last
relay lockout malfunction occurred on March 18,
1983.
The seven writeups on the engine-driven generators concerned the generators*
tripping off line.
On August 1, 1982, the right generator tripped and was reset by the
flightcrew; on August 2, the right generator’s voltage regulator was replaced. The six left
generator malfunctions occurred between December 28, 1982, and January 4, 1983. On
January 4, 1983, the left generator control panel was removed. and replaced and
thereafter the generator operated without further problems.
A shop check of the
removed control panel disclosed a bad solder connection between circuits on the under
frequency protection printed circuit board.
National Transportation Safety Board Accident Report AAR-80-13.
Air Canada Engineering Report No. 920-754-3, December 10,
”
’
-
I
Between May 7, 1983, and June 2,
1983, the flightcrews logged 38 APU
generator malfunctions;
during this period there were no engine-driven generator
.
malfunctions logged. Thirty-four writeups indicated that the APU generator would not
come on line; four indicated that the generator had tripped from the line. The
examination of the flight log showed that corrective action had been taken for these
entries.
Except in an electrical emergency, the
generator is not used in flight.
Flight
flightcrew did not try to start the APU generator during the flight from
Dallas to Cincinnati.
Between September 1,
1980, and June 2, 1983, the cabin logbook entries
indicated only minor deficiencies in the toilet pump flushing system and routine flush
pump and flush motor changes; all were signed off properly with no out-of-the-ordinary
repairs having been made.
The last flush system component change was made on May 4,
1983; the pump assembly was replaced and the repair signed off by a mechanic and an
inspector.
On May 2, 1983, during a scheduled major maintenance check, an unscheduled
work card -- No. 150 -- contained the following writeup,
at bottom of
pressure bulkhead in rear cargo (compartment) soaked with toilet detergent liquid,
investigate
The item was signed off, as follows, checked and
tightened. Also, insulation replaced where
1.6.2
Passenger Cabin Modification
During June 1982, Air Canada refurbished the airplane’s passenger cabin. The
right aft lavatory was removed and replaced with a clothing stowage area.
Overhead
luggage bins were installed and the cabin walls and ceilings were replaced.
The
modification was performed using an assembly kit manufactured by the Heath Tecna
Corporation, Kent, Washington, in accordance with the provision of Supplemental Type
Certificate
No. 1429 issued by the FAA Northwest Regional Headquarters,
Seattle, Washington, on February 2, 1982.
Since the Heath Tecna assembly kit was designed to be used on all DC-g-32
airplanes, Heath Tecna had to apply for and received an STC. The materials, drawings,
and plans contained in the assembly kit constituted a major overhaul of the airplane’s
interior.
Since the DC-g-32 airplanes had been certificated before May 1, 1972, the
manufacturer had to demonstrate that materials met the flammability standards of 14
CFR 25.853 (a)
as amended on May 1, 1972, in order to receive an STC; these standards
still apply. (See appendix
The flammability tests were conducted in accordance with
prescribed FAA standards, and all materials in the kit met the flammability standards. In
addition, all wiring used in the kit met MIL W-81044 specifications previously approved by
the FAA.
At the time of the accident, the airplane’s seat cushion material was
polyurethane foam, the window panes were transparent acrylic sheet, the interior cabin
side walls were made of acrylonite butadiene styrene
plastic sheet, the ceiling
panels were made of composite nomex honeycomb core with fiberglass facing materials,
and the cabin sidewalls and ceiling panel facings were decorative vinyl laminate with
Tedlar facings.
-
-
”
’at
-
’
-
1.7
-12-
.
Meteorological Information
The 1700 National Weather
Service
surface analysis for June 2, 1983,
showed a west to east warm front in central Kentucky.
At 2000, the analysis showed a
warm front extending from southern Ohio through western Kentucky with associated rain
and rainshowers.
Weather radar data from the Cincinnati Airport for 1830 and for 1930
indicated weather echoes containing rainshowers.
These echoes were located south
through west of Cincinnati Airport and extended out to about 100 nmi. The maximum
echo tops were reported at 14,000 feet at 1830; at 1930, the tops were reported at 13,000
feet. According to the NWS, light rain began at the airport at 1734 and ended at 2024.
Between 1900 and 2000, a trace (less than 0.01 inch) of rain was measured.
Surface weather observations for the Cincinnati Airport were as follows, for
the times indicated:
1850
2,500 feet scattered, measured ceiling 8,000 feet overcast;
visibility --
12 miles, light rain; temperature --
F; wind -- 7 knots; altimeter setting -- 30.04
1930 2,500 feet scattered, estimated ceiling 8,000 feet overcast;
visibility -- 12 miles, light rain; temperature --
F; --
55’ F; wind
5 knots; altimeter setting -- 30.03
1.8
Aids to Navigation
Not applicable.
1.9
Communications
There were no known radio communications difficulties.
1.10
Information
The Greater Cincinnati International Airport, elevation 891 feet, is located 9
miles southwest of Cincinnati, Ohio, in Covington, Kentucky. The airport is certificated
for commercial operations in accordance with 14 CFR 139, Subpart D.
The landing area consists of three runways:
18/36, 9R/27L,
and
Runway 27L is 7,800 feet long and 150 feet wide, and has a grooved concrete and asphalt
surface.
Runway 27L has high intensity runway edge lights centerline lights, a
medium intensity approach light system with runway alignment indicator lights
and visual approach slope indicator The touchdown zone elevation is 875 feet.
Runway
is served by an ILS approach.
Standiford Field, elevation 497 feet, is 5 miles south of Louisville, Kentucky.
The airport is certificated for commercial operations in accordance with 14 CFR 139,
Subpart D. The landing area consists of two runways:
1-19 and 11-29. Runway 1-19 is
7,800 feet long and 150 feet wide, and has a concrete surface, HIRL, and an approach
light system.
Runway 11-29 is 7,429 feet long and 150 feet wide, and has an asphalt
c
_
’s
’s
’
-
.3
surface, HIRL, and runway end indicator lights Runway 29 has an approach‘light
.
system; runway 11 has a VASI, but no approach lights.
Runways 11 and 19 are served by
.
an ILS approach; runway 11 is served by a
(back course) approach.
1.11
Flight Recorders
The airplane was equipped with a Leigh VDR-2 digital flight data recorder
(DFDR), serial No. 127, and a Fairchild A-100 cockpit voice recorder
serial
No. 1613. Both recorders were removed from the airplane after the accident. The CVR
was brought to the Safety Board’s Audio Laboratory for processing and readout.
Since ‘the
Safety Board’s Washington laboratory is not equipped to readout the Leigh DFDR, the
readout was performed at the Flight Research Laboratory, National Research Council
(NRC), Ottawa, Ontario, Canada, and was observed by Safety Board personnel.-.
Cockpit Voice Recorder.-The CVR casing was damaged by fire and smoke;
however, the crash-proof enclosure protected the tape and the quality of the recording
was excellent. A tape was transcribed beginning at
and ending at when
the CVR ceased operating.
Using the time signal recorded on the
Indianapolis
tape as a basis for comparison, the CVR tape timing was accurate to the second.
(See appendix
The entire CVR tape was examined for sounds of electrical arcing or other
events which might-be associated with the accident.
About 10 minutes into the tape, at
a sound resembling that of electrical arcing was recorded. The sound was
repeated at
1848:15, 1851:03, 1851:05, 1851:14, 1851:42, 1859:59,
and 1900 .
The
crewmembers testified that they did not hear arcing sounds at these times.
The
content of the first two electrical arcing sounds differed from
those which followed. The early sounds were impulse-type and contained a broad band of
frequencies resembling radio static. All of the later arcing sounds contained a 400 Hz
component with harmonics extending through the frequency range of the recorder.
At
an electrical pulse was recorded simultaneously on both the
captain’s and first officer’s radio channels. The pulse, which lasted about 7 milliseconds,
occurred about the time the captain said that the left
bus was lost. Signals from the
radio channels are taken from the captain’s and first officer’s audio selector panels.
Flight Data Recorder.-The recording medium of the Leigh DFDR is a
l/a-inch,
ontinuous
loop, ‘I-track magnetic tape.
A total of 33.5 hours of data
encompassing 76 airplane performance parameters are recorded. The recorder, which was
not damaged, was opened at the NRC flight recorder laboratory; the tape was removed
and then wound on a standard computer-tape reel. The accident flight was identified both
by the recorded flight number and by tracing the altitude and heading time histories from
the takeoff at Dallas. The data showed that the recorder stopped operating 1 hour 42
minutes into the flight while the airplane was at FL 330.
The DFDR recording contained several anomalies that took the form of signal
spikes or data losses in a number of recorded parameters. These anomalies were used to
establish a correlation between the CVR and DFDR times.
Since the. DFDR did not
contain microphone keying information, it was necessary to identify events that were
to both
DFDR anomalies
signals. which
had been recorded on the CVR channels.
The time
-
*-
-14-
,
increments between the specific events, recorded on the CVR and DFDR, were
compared and a correlation established.
From this information, it was determined that
the DFDR also stopped recording at
Wreckage and Impact Information
Both engines and their associated cowlings were intact, undamaged, and
showed no evidence of exposure to abnormal heat or fire. There was no evidence of oil or
fuel leakage, and the main engine fuel supply system did not leak when pressure tested.
The engine fire extinguisher bottles had been discharged.
The APU was intact and was not damaged. The exterior of the APU and
surrounding compartment were free of soot and other fire damage. The APU enclosure
within the aft accessory compartment was not damaged by fire; however, the enclosure
was coated slightly with soot.
Visual inspection and tests of the hydraulic and fuel systems revealed that
neither system contributed to either the initiation or propagation of the fire. The
empennage and wings were not damaged by either fire or heat. The leading edge slats and
trailing edge flaps were fully extended. The nose gear was extended and locked. Except
for the right axle where the splash guard had been cut away in order to tow the airplane
from the runway, the nose gear was not damaged.
The nose wheel tires were inflated.
Both main landing gears were extended and locked. Except for the support bracket on the
left main gear, which was bent and twisted slightly aft, the main landing gear was not
damaged. However, all four main wheel tires had blown on landing.
1.12.1
Fuselage
The cabin area of the upper fuselage down to below the level of the cabin
windows was damaged heavily by fire; below that level the fuselage was relatively intact.
The majority of the cabin windows were either missing or had partially melted out. (See
figures 3 and
Forward of the aft pressure bulkhead, the upper areas of the cabin
windows were discolored and burned away in several locations. The fuselage skin above
the left aft lavatory was intact, but a rectangular area corresponding closely to the shape
of the lavatory had been discolored to dark brown and a large area of paint had been
burned away. The rectangular area began above the engine pylon and extended to the top
of the fuselage. (See figures 3 and 4.)
Forward of the aft lavatory, between fuselage station
929 and FS 758, the
top of the fuselage was damaged heavily by heat and had been burned away down to the
top of the cabin windows. Between FS 758 and FS 484, the fuselage upper skin was intact
but was buckled and discolored by heat at the very top between the 11 o’clock and 2
o’clock positions (aft looking forward).
Between FS 484 forward to the cabin entry door
the upper skin of the fuselage had burned away partially. Except for some
sooting around the edges of the left and right forward cabin entrance doors, the fuselage
forward of FS 200 was intact with no apparent heat or fire damage.
There were thick soot deposits along the lower side of the fuselage beginning
at the cabin air outflow valve at FS 945 and from around the access
in the lavatory
service panel at FS 965. The soot pattern trailed rearward along the airplane’s side,
Figure S.-Left side of the airplane.
Figure 4.-Left side of the airplane, rear view, depicting rectangular
burn pattern and Soot trail. .
.
_
-
-17-
.
.
Figure S.-Soot trail deposits on left side of fuselage
at lavatory. service panel and outflow valve.
Figure service with access door open.
.
_
ENTRANCE
DOOR
.
\
\ \
OXYGEN MASK
COMPARTMENT
ASHTRAY
SINK BASIN
COLD AIR
OUTLET
SANITARY
NAPKINS
BAG
DISPOSAL
‘SICK BAG
CONTAINER
of aff
lavatory5 viewed
aft.
Figure lavatory viewed toward aft and outboard wall.
Lavatory sink door is open exposing trash receptacle:
Toilet shroud has been removed showing the top of the
flush motor between the commode and forward wall of
amenities section.
’F.
-21-
The aft lavatory was damaged extensively by fire and heat. Most of the
lavatory’s interior walls had been burned away.
The entry door, which had been kept
closed, was destroyed except within 16 inches of the floor. The aft pressure bulkhead and
fuselage skin which comprised the outermost portion of the lavatory enclosure were
intact, but were buckled and discolored to dark brown.
The stainless steel sink and the section of the vanity frame supporting the sink
were intact against the aft lavatory wall.
The plastic door on the front of the sink portion
of the vanity was burned away.
The cold air outlet nozzle mounted in the door was
recovered in the debris and was in the closed position.
Except for the top of the trash chute, which had burned away at the point
where it attached to the top of the sink at the waste disposal door, the trash chute and
trash container behind and below the sink were intact. The aft side of the container was
scorched, but not burned, and the paint on the inside surface of the container was
blistered. Paper trash in the container was scorched but was not burned. The aluminum
shelf below the container was intact; it was covered with debris, but exhibited little
evidence of heat damage. The lavatory floor in this area was intact, but was covered with
damp debris.
Included in the debris were a plastic vial and a paper maintenance tag,
neither of which was burned. These items were located on the floor under the sink area.
A
1301 automatic fire extinguisher was mounted below the sink basin,
and it had discharged automatically into the trash chute. The extinguisher is designed to
discharge
gas through one, or both, of two heat activated nozzles. One nozzle
discharges directly into the open area below the sink and the other into the trash chute.
At temperatures in excess of
the nozzle tips will melt causing the to be
discharged under pressure.
The amenities section of the vanity extends from the sink to the lavatory
outboard wall and from the commode to the lavatory aft wall. The section contains the
oxygen mask compartment and containers or dispensers for paper towels, toilet paper,
sanitary napkins, and sick bags. (See figure
The amenities section had been damaged extensively by fire and was almost
completely disintegrated along the outboard corner.
The oxygen mask compartment was
damaged severely and parts of.the compartment and its cover door were burned away.
The lowest level of the amenities
section between the amenities
compartments and the aluminum bottom shelf was a void space through which was routed
the wiring for the
flush pump motor; the aluminum oxygen line to the oxygen
mask compartment; the aluminum cold air supply line to the adjustable nozzle outlet
below the sink; and a stainless steel pipe from the sink.
The outboard top of this area had
burned away.
The outboard corner of the aluminum bottom shelf had partially melted
away; the remainder of the shelf was intact.
The lavatory aft wall below the bottom shelf
was intact and was not burned.
Within this portion of the amenities section, the stainless steel drain pipe was
intact. The oxygen line had been partially consumed where it entered the oxygen mask
compartment and where it exited the amenities section; the remainder of the line within
the amenities section was intact. The cold air supply line was intact from the divider wall
between
sink and
pressure bulkhead and the outboard wall of the lavatory, In this area, a 1.5-inch-long
elliptical-shaped hole was melted through the top of the line. The aluminum alloy used in
the cold air supply line melts between 1,200’ F. The remaining outboard end
of the cold air supply line was scorched and partially melted away about 2.5 inches beyond
the elliptical hole.
There was extensive heat damage to the vanity wall and floor in the
area adjacent to the elliptical hole.
The plastic top and shroud of the commode were burned away. Except for
some burning around its upper edges, the commode’s fiberglass waste tank was relatively
intact, was partially filled with debris, and contained about 6 to 7 inches of water. The
flush pump and flush pump motor assemblies were recovered in several pieces among the
debris in the waste tank.
The waste tank in the commode is serviced by a stainless steel flush and fill
line which runs from the tank to its terminus in the lavatory service panel. The stainless
steel pipe was intact, however, the flex hose and joints which connect the pipe to the
waste tank had burned away.
The plastic ball in the flush and fill pipe check valve had
burned away and the brazed joint between the two portions of the check valve housing was
partially melted. Soot and tar deposits were found on the lavatory service door directly
opposite the flush and fill line outlet, and rivets on the pipe connector to the service panel
had been melted.
The flush motor was found attached to its mounting flange which was attached
to the pump well.
The flush motor assembly was found in approximately its normal
position in the waste tank next to, but detached from, the flush pump and filter assembly.
The motor housing was completely melted.
Large concentrations of molten metal were
found on the forward side of the mounting flange facing the toilet bowl, and on the aft
side where the
electrical leads enter the motor housing. A portion of the wiring
harness which supplies power to the flush timer, flush motor, and flush button was
encased in the molten metal on the aft side of the motor. The rotor section of the flush
motor was encased in molten metal above-the field windings; however, the rotor did not
appear to have been damaged extensively by heat. (See figure
The flush pump and filter assembly found in the waste tank was partially
intact. The filter and pump mechanisms, which apparently had been located below the
water line in the waste tank, were in good condition.
Above the water line, the plastic
housing for the pump and filter was melted away.
This housing also encloses the filter
gear train and is part of the assembly used to mount the pump and filter to the pump well.
The exposed shaft and gears showed heat damage.
The gears, which make up the filter
gear train that connects the motor to the pump and filter, were found lying in the waste
tank. The plastic mounting ring, which mates the pump and filter housing to the pump
well, was found intact among the debris in the waste tank. There was some slight melting
of the top surface of the mounting ring.
All of the components of the flush motor and
pump assembly were removed for further examination.
When the flush motor and its associated mounting were removed from the
toilet tank, two wires in the motor harness and all of the wires in the power harness broke
just aft of the motor housing due to brittleness.
To facilitate removal of the flush motor
assembly, the remaining wires of the motor harness were cut a few inches aft of the
motor housing and tagged for identification.
ASM Metals Handbook, Vol. 2, Edition 9.
’s
_
-
The flush timer was found intact, mounted in its normal position, on the
inboard interior cabinet wall below and to the left of the sink. The timer’s two electrical
connectors and their respective wiring harnesses were connected to the timer. The timer
and connectors showed evidence of external heat and smoke damage only. The connectors
were removed and all mating connections were examined. No damage was observed at
any of the connections. The timer was removed for further examination.
The wiring harness that supplies 3-phase power from the timer to the toilet
flush motor was examined. The harness was undamaged between the timer and a point
midway into the lower amenities portion of the vanity.
The insulation had melted away
from the wires on the wiring harness section between the midway point and the lightening
hole
where the harness leaves the vanity and connects to the motor. The exposed
wires% this section were brittle.
The wiring harness that supplies
power from the ground service bus to the
timer and flush button was found partially intact. The harness was undamaged in the area
from the timer outboard to the lower amenities portion of the vanity. From this point,
continuing outboard and forward through the lightening hole in the vanity, the wire
insulation and the wires were brittle. The harness was encased in molten metal from the
aft end of the motor housing, across the width of the housing, to the forward end of the
housing. The harness was not identifiable from a point just forward of the motor housing
to slightly aft of where the harness leaves the lavatory forward bulkhead through an
overhead lightening hole. From the overhead lightening hole, the harness was routed
outboard and down to the cabin floor.
Varying degrees of heat damage were observed on
the harness in this area. The remainder of the harness, from a point just below the cabin
floor forward along the left side of the airplane to the circuit breakers on the electrical
power control panel, was undamaged.
No wiring from this harness to the flush button
could be identified.
The lightening hole between the amenities section and toilet section through
which the flush motor’s power harness passes was examined. The nylon alligator grommet
covering the
surface was not found; however, this entire area was damaged severely
by heat and fire. The bracket supporting the harness! nylon cable clamp was partially
melted away with only the portion of the bracket that was riveted to the structure
remaining.
An electrical continuity check was made of the harness using a volt-ohm
meter.
Because the insulation had burned off the wires in certain areas causing short
circuits, the entire harness could not be tested.
The harness was cut at a point 5 feet
above the cabin floor, just forward of the aft lavatory forward bulkhead. Continuity was
observed from this point to the circuit breakers.
There was no evidence of line-to-line or
line-to-ground short circuits.
Electrical splices were found on the power harness just aft of the flush motor
housing. The
four of them. The remaining splices could not be tested because there was not sufficient
wire protruding from the splices.
A hole resulting from the removal of non-load carrying metal in airplane members for
the purpose of weight reduction.
.
’
’
_
l
,
.
.
. .
.
The power harness routing in this airplane differed from the McDonnell
Douglas DC-9 installation drawings.
These drawings indicate that the harness is routed
from the lightening hole in the vanity, across the inboard side of the flush motor housing,
and around the forward side of the housing.
The harness is shown secured
support
clips and cable clamps around the pump well.
The harness in this airplane was routed
across the outboard side of the motor housing.
No support clips or cable clamps were
found around the pump well. It could not be determined whether or not vibration induced
insulation damage occurred at this point.
The wiring harness associated with the aft attendant communications panel
was examined.
A number of electrical wire splices were found just above the standpipe
feed-through located near the inboard end of the vanity. This harness and the attendant
panel were removed for further examination.
Electrical components normally located in the lavatory overhead area were
found among the debris recovered from the lavatory floor around the waste tank. All
components were heavily damaged by fire.
The recovered components included portions
of the upper and lower mirror light assembly, the circuit breakers and transformer to the
aft reading lights, the razor convertor, and the control transformer for the aft miscel-
laneous lights.
All the overhead wiring from the aft pressure bulkhead forward to the cockpit
was severely damaged by fire and heat.
The wiring that penetrated the aft pressure
bulkhead was spliced just forward of the bulkhead.
All of the splices observed in the area
of the aft pressure bulkhead and in the area of the aft lavatory were accomplished during
repairs made to the airplane after the September 17, 1979, aft pressure bulkhead
separation at Logan International Airport, and the splices were made in accordance
service sketch 2958. None of the wire splices in these areas showed evidence of arcing or
shorting. Several samples of electrical wiring, all of which were spliced, were removed
from the forward side of the pressure bulkhead for further examination. All of the wire
splices removed and examined appeared to exhibit the same degree of extreme heat
damage.
The insulation covering was missing from all of the splices; however, no
evidence of either electrical arcing or shorting was observed.
1.12.4
Aft Accessory Compartments and Cargo Compartments
The internal fuselage of the aft accessory compartment behind the aft
pressure bulkhead was intact with little evidence of heat damage. The insulation on the
aft side of the bulkhead was intact but discolored. Except for the buckling and
discoloration that was noted in the area above the aft lavatory% vanity, the aft pressure
bulkhead was intact. The systems, lines, and wiring were intact except for some slight
heat damage to the wiring nearest the lavatory.
The forward cargo compartment was fully intact, and there was no evidence of
either fire or heat damage.
The rear cargo compartment also was intact. The fire and
heat damage in this compartment was concentrated in the area below the aft lavatory.
The under floor blanket insulation along the aft. tunnel area of the rear cargo
compartment was scorched from the aft pressure bulkhead forward to about FS 945 and
from the airplane’s centerline outboard to the fuselage skin. The heaviest scorching was
under the aft lavatory at the point where the lavatory’s aluminum vent tube was routed.
This tube, which vents air overboard from the. lavatory, begins near the commode and
beneath the toilet shroud and ends at a venturi- in the lavatory service door. It enters the
tunnel just aft of FS 980 and is routed forward to FS 965 and then to the venturi.
’
Between FS 980 and FS 965 the tube is routed below the three generator feeder cable
bundles. The vent tube had melted away to within 6 inches of its terminus in the lavatory
service panel. The hydraulic, fuel, and pneumatic lines routed through this area were
intact with no evidence of leakage.
The nylon conduits encasing the APU and the left and right generator feeder
cables from the aft pressure bulkhead
1019) to about FS 965 were melted away and
the insulation on the exposed cables was scorched.
The most intense scorching occurred
between about 2 inches forward to 8 inches aft of FS 980. The generator feeder cables
are routed through two lightening holes in a floor beam located at FS 980. The APU and
right generator feeder cables pass through the inboard lightening holes; the left
generator’s three feeder cables pass through the outboard holes. A small notch, typical of
the type of damage resulting from electrical arcing, was found in the outboard lower edge
of the inboard lightening hole adjacent to one of the right generator feeder cables. The
frame of the outboard lightening hole was burned away from around the bottom of the
hole.
The nylon support clamps for the feeder cable bundles were missing at both
lightening holes. The clamp screws for the nylon support clamps were attached at the
inboard hole, but were discolored by heat.
At the outboard hole, the clamp screws were
missing and the area where they attached was damaged by heat and partially burned away.
At FS 965, the nylon support
for the APU feeder cables were intact
while those for the right generator feeder cables were only partially intact. The support
clamp for the left generator feeder cables was missing; however, the clamp screw was
.
attached at the lightening hole.
There was some heat damage, but no evidence of arcing
on either lightening hole at FS 965.
The support clamps were intact and the generator
feeder cables were supported properly at the remaining lightening holes observed.
The examination of the feeder cables of the two engine-driven
the APU
generators showed that each had been damaged by intense heat in the area between FS
965 and FS 996. The feeder cables of the right generator showed evidence of arcing near
FS 980. The nylon conduits containing the generator feeder cables are semi-rigid,
structures.
During the investigation, the support clamps in the lightening holes at FS 980
were removed from a sister DC-9-32. The conduits remained in place and did not contact
the surfaces of either lightening hole.
The feeder cables of the engine-driven and the APU generators were
disconnected at the generator relays and at the engine and APU
connectors.
When tested, each line showed continuity and no line-to-line or line-to-ground short
circuits were observed.
Eight-foot-long sections of the three generator feeder cable assemblies were
cut out and removed for closer examination.
Each feeder cable assembly exhibited an
area about 2.5 feet long wherein its nylon conduit had melted away and the insulation
within this melted area was brittle and charred heavily.
Examination of the left generator cable showed that the
and C-phase lines
had areas wherein the insulation had chafed. A metal globule was found on the exposed
wire strands in this area. Another chafed area was in the C-phase line about inch aft
of the metal globule. There was some melting of the wire strands in this area; however,
no evidence of arcing was found.
_
-
l
.
-27-
A small area of chafed insulation was found on the phase B line of the right
generator feeder cable bundle. A metal globule, similar to that noted on the left
generator lines, was found on the exposed wires.
The areas where the metal globules were found on the left and right generator
lines correspond to where the lines appeared to have contacted the floor beam structure
under the aft lavatory at FS 980.
The APU feeder lines showed no evidence of chafed
insulation or electrical arcing.
The examination of the airplane’s electrical wiring included the components in
the electrical and electronics compartment located below the cockpit. This compartment
includes the components used for
and bus power distribution and fault protection,
and the electronic components used
communications, navigation, and flight control.
There was light to moderate heat, water, and firefighting foam damage observed in the
compartment; however, the wiring harnesses showed little heat damage.
All components
were intact and were mounted properly in their respective racks. The batteries, which
had been disconnected by the firemen, were intact. The following were removed for
further testing: the static inverter; the voltage regulators of the three generators; the
left and right generator control panels; and the
bus control panel.
1.12.5
Controls and Instruments
The readings of the cockpit instruments,
the
of the cockpit
controls, and the positioning of the switches in the cockpit were’also documented during
the investigation.
With regard to the air conditioning and pressurization packs, the ram
air switch was off, the right air conditioning supply switch was off, and the left air
conditioning supply switch was in the HP (high pressure) bleed position. The left air
conditioning supply switch was reported originally to have been in the off position, and
investigators could not determine if the switch had been moved.
1.13
and Pathological Information
Blood samples were taken from the 18 surviving and 23 deceased passengers
and were analyzed by the FAA’s Civil Aeromedical Institute, Oklahoma City, Oklahoma,
for carbon monoxide, cyanide, fluorides, and ethyl alcohol. The results of the analyses
indicated that the deceased had elevated carbon monoxide levels ranging from 20 to 63
percent saturation; the threshold for carbon monoxide in the blood at which incapacitation
occurs is between 40 and 50 percent saturation. The cyanide levels found in the blood
samples of the deceased ranged from a low of 0.8 to a high of 5.12 micrograms/ml; the
toxic level for cyanide in the blood at which incapacitation occurs is between 0.5 and 0.7
micrograms/ml.
The fluoride levels ranged from 410 micrograms/100 ml to 63
micrograms/100 ml; however, the significance of these fluoride levels is unknown.
Alcohol levels on three of the deceased were in excess of 0.10 percent concentration.
Blood samples were taken from the survivors about 2 to 3 hours after the
accident.
The concentrations of carbon monoxide and cyanide found in the survivors’
blood samples were below 0.10 percent concentration and
micrograms/ml,
respectively. Fourteen of the 18 survivors’ blood samples tested negative for alcohol; the
other 4 samples tested below 0.10 percent concentration.
With regard to the blood
alcohol levels, since the blood samples were taken 2 to 3 hours after the accident, these
values may be low. Blood alcohol levels decrease at about 0.015 percent per hour after
alcohol intake has ceased..
.
but
”
’
Autopsies were performed on five bodies under the direction of the Boone
County Coroner, and an additional five under contract for Air Canada. No evidence of
antemortem impact injuries was discovered during these examinations.
1.14
Fire
Response
Although the fire on board Flight 797 began in flight, no one saw flames in the
cabin until after the flight had landed and the survivors had left the airplane. The last
passengers to depart the airplane through the left and right
emergency exits
stated that they saw flames immediately after stepping onto the wing. The firefighter on
scene commander stated that some of his men went to assist the passengers down from
.
the wing and that, at that time, he saw flames in the cabin.
The crash-fire-rescue vehicles entered runway 27L at its approach after Flight
797 landed and followed behind the flight until the airplane was stopped. According to the
fuel gauge readings noted during the cockpit documentation, the center wing tank was
empty and there were 6,200 pounds and 6,050 pounds of jet-A fuel in the left and right
wing main tanks, respectively. The airplane’s fuel tanks did not rupture and the jet-A fuel
was not involved in the fire.
Flight 797 came to a stop about 1920, and 7 airport crash-fire-rescue vehicles
containing 13 airport firefighters were positioned at the
(See table Upon
arriving at the airplane, the firemen saw heavy smoke out of the exits
and front doors. About 1921, as surviving passengers
crewmembers were departing
the airplane, the firefighters initiated an exterior attack on the fire.
Foam was
discharged from the turrets on the firetrucks onto the top of the airplane% fuselage and
on the ground beneath it in order to cool the interior of the airplane and to provide a foam
blanket in case of a fuel spill.
Other firefighters assisted passengers at the escape slides
and helped passengers off the wing to the ground.
. When the on scene commander ordered an interior attack on the fire to be
made for rescue purposes and to extinguish the fire, passengers’ were still leaving the
airplane through the left forward cabin door and
window exits. The first interior
attack was made through the left aft window for several reasons.
According to the on
scene commander who executed the order, he believed that most people would try to exit
the airplane through the left forward door; therefore, he did not want to block that exit
with a ladder and hose. Also, with the escape chute deployed, it would have been difficult
for the firefighters to enter at that door with their protective equipment and hoses. The
on scene commander testified that it would have been possible,
it would have slowed
us
The on scene commander also testified that he wanted to enter the airplane
through the
window exits, “because we wanted to get in between the passengers
and the fire to make their chances (to escape) better.”
Within minutes of arriving at the airplane and after the passengers were off of
the left wing, two firefighters mounted the left wing carrying a 1.5-inch handline, opened
the left aft
emergency exit, and applied foam into the cabin. The firefighters
were wearing proximity suits with self-contained breathing apparatus; however, they were
not wearing the proximity
protective hoods because the hoods did not fit over their
After applying the foam into the cabin, they attempted to enter it
through
exit but were driven back by the intense smoke and heat.
According
to one of the firefighters, he did not see any flames during this attempt to enter the
5
Table l.--Responding Airport Crash Fire Rescue Equipment
Vehicles/
Unit No.
Agent
Capacity
(gallons)
Discharge
Rate
Crash truck
Unit 907
.
500 AFFF*
*
Unit 913
3,000 *
750
500 protein
Quick Reaction
Vehicles
Unit 967
100 AFFF
450 pounds
Purple K. ***
Engine Companies
Unit 951
1,000
Ladder Companies
Unit 960
1,500
Ambulances
Unit 964
Rescue Squads
Unit 980
n/a
*Water
. .
**Aqueous film forming foam
***Dry chemical extinguishing agent
+ Indicates more than the cited amount was used.
Quantities
Used
allons)
3,000
+
3,000
+
0
0
1,000
300
.
*
-
‘..
”
.
.
Air Canada Publication 356 contains both pictures and descriptions of the fire
extinguishers used on their airplanes; it also depicts where each fire extinguisher is
located on each airplane operated by the company.
Publication 356 explains and depicts
how each fire extinguisher operates and what type extinguisher should be used to fight
different types of fires.
With regard to a lavatory fire, Publication 356 states, in part,
“Execute flame knockdown by repetitive discharges of a carbon dioxide
or dry
chemical extinguisher.”
The publication also contains directions for the most effective
use of each type of extinguisher. According to the manual, the user of a extinguisher
should,
“AIM the gas at the outside edge (of the fire) and then in a circling fashion
towards the center.”
All Air Canada flight attendants receive “hands- on” training in the use of all
fire extinguishers during initial
training: In addition,-during initial training
they are required to extinguish an actual fire.
Publication 356 also states, in part, that if excessive smoke and fumes are
present in the cabin, the flight attendants should “relocate passengers away from the area
of severe smoke and fumes if possible.” The passengers had been moved forward in the
cabin, and no passenger was seated farther aft than row 12. The two passengers in seats
Nos. and refused to move forward because their seats were next to the right
forward exit window. (See figure 12.)
Once the passengers had been repositioned and the cabin air vents opened and
directed aft, the smoke appeared to lessen, but shortly thereafter the smoke began to
increase rapidly. Several passengers stated
the cessation of airflow from the vents
coincided with the increase in the smoke. Other passengers stated that it occurred at the
beginning of the descent or sometime shortly after the airplane began descending.
Air Canada emergency procedures state that the oxygen masks should not be
deployed below 10,000 feet as a means of avoiding smoke inhalation. Below 10,000. feet
less than 1 liter per minute is being supplied through the mask, and therefore, due to the
design of the mask and the low altitude, the user is merely breathing ambient cabin air.
The company procedures also state,
fire conditions exist, dropping the masks and
pressurizing the oxygen manifold may contribute to combustion.” The procedures further
note, “If loss of cabin pressure has caused the masks to drop. . .
the passengers should
remove them as soon as practical once the cabin pressure altitude drops below 13,000
feet.
The flight attendants designated several male passengers to open the
exit windows after the airplane landed and stopped. None of those designated could recall
whether the attendant had given them specific directions as to how the exits were to be
opened. However, nearby passengers recalled hearing a flight attendant describing the
operation of the
emergency exit windows.
Three of the four emergency
exit windows were opened by passengers, and none encountered difficulties in
operating and removing the window exits.
During descent, the cabin filled with black, acrid smoke from the ceiling down
to about knee level. Passenger and flight attendant testimony and statements indicated
that all of the surviving passengers had covered their faces with either wet towels
distributed by the flight attendants or articles of clothing. They all attempted to breath
as shallowly as possible, and all reported that the smoke hurt their noses, throats, and
chests and caused their eyes to water.
By the time the airplane- landed, they could not see
I
come
”
Land
did
”
”
joined by the No. 2 attendant, and they both moved forward rechecking seatbelts
comforting passengers.
When they reached the forward cabin area, the No. 2 flight
attendant sat down in row 3C and remained there until the airplane landed. The No. 3
flight attendant moved aft again.
She sat down in an aisle seat in rows 7, 8, or 9 and
remained in that seat until the airplane landed.
While seated, she shouted
instructions to the passengers before the airplane landed. Several passengers said that
they heard these instructions.
After the airplane landed and stopped, the flight attendant in charge opened
the left forward cabin door, inflated the slide, and sent a passenger seated on the jump
seat down the slide.
He then positioned himself in the doorway and shouted for the
passengers to,
this
One of the passengers recalled hearing that order.
Another passenger testified that by this time, given the conditions in the cabin, it was
doubtful if anyone could draw sufficient breath to shout loud enough -to be heard at any
distance.
With regard to the flight attendant’s duties during a
Evacuation With
Warning,
Publication 356 states,
“Flight attendants should do all possible to evacuate
everyone, but are not obliged to risk their own lives.” The flight attendant in charge
testified that he stayed in the left cabin entrance doorway until no more passengers were
coming. At that time, the heat was becoming too intense to remain, and he exited the
.
The No. 2 flight attendant went forward after the airplane stopped. She saw
the attendant in charge open the forward door and deploy the slide. She saw a male
passenger exit through the door, and she followed him out of the airplane. Thereafter, she
helped and supervised other passengers as they left the airplane.
The No. 3 flight
attendant got to her feet after the airplane stopped. The smoke was so thick that she
could not see. She testified that she
not think to go back to the exits, so
she felt her way forward
She went to the right forward cabin entrance door, opened-the
door, and inflated the slide. She stood in the doorway, yelled,
this waited 3
to 4 seconds, and then exited the airplane down the slide. She waited a couple of seconds
at the bottom of the slide and when no one came, she ran around to the left side of the
airplane and began to assist the other crewmembers in rendering aid to the survivors.
Seven passengers and two flight attendants exited the airplane through the left
forward cabin entrance door and slide; one flight attendant exited through the right
forward door and slide; four passengers exited through the right forward
emergency exit window; one passenger exited through the right aft emergency
exit window; and six passengers exited through the left forward
emergency exit
window.
The
.three overwing
exit windows were opened by designated passengers. The
smoke in the cabin was reportedly so thick that most of the passengers had to get to the
exits by using the seatbacks to feel their way along the aisle. None of the passengers
noticed if the emergency lights were illuminated. Several passengers said that, when they
either bent forward or got on their hands and knees, they were able to breath and see’ a
little better, but it was not much of an improvement.
One of the passengers who used an
emergency window exit said that she was able to locate it when she saw a very
dim glow of light coming through the aperture.
Another stated that she was able to
locate the
emergency exit window when she felt a slight draft on the back of her
,
knees.
During the evacuation, passengers in the seats 2-B, 2-E,
3-C, 3-E,
and 8-C exited through the left forward cabin door; passengers in seats 9-E,
and 11-E exited through the left forward window exit; passengers in
.
_
’
_
-
.
seats
10-E, 11-C,
12-D, and 12-E exited through the right forward window exit;
.
and the passenger in seat 9-C exited through the right aft window exit.
l
Except for two fatalities found in the aisle at rows 14 and 16, the majority of
the fatalities were found either in the aisle or seated in rows 2 through 9. (See figure 12.)
,
The fatalities at rows 14 and 16 had been seated in seats 8-B and 9-B, respectively.
1.16
Tests and Research
Jnvestigation
LaboFatory Tests
The following items from the aft lavatory of Flight 797 were delivered to the
.FBI laboratory for analysis:
samples of waste tank water; fiberglass insulation from the
aft lavatory; a plastic vial and tag recovered from the lavatory floor; an aluminum shelf;
fiberglass flooring; and soot deposits from the inside of the lavatory service panel access
door.
The results of the examination were as follows:
No flammable
were identified on the items listed The source of the spots on the fiberglass
flooring could not be determined.
The soot deposits contained residues which were
characteristic of a phenolic residue, resulting from the burning of phenolic resins such as
those contained in the cabin and lavatory walls and other materials.
lJ6.2
Electrical System Components
The electrical system components removed from the forward electronic
compartment beneath the cockpit. were tested under Safety Board supervision at the
Westinghouse Electric Corporation, Lima, Ohio.
Functional testing of the a.c. bus control panel, and the voltage regulators and
generator control panels of the APU and the left and right engine-driven generators
showed that these units. were operational.
Inspection of the APU generator control panel revealed that connectors on the
printed circuit board had corroded and that the electronic components on one of the
printed circuit boards had been damaged by water and foam. The Safety Board concluded
that this was damage incurred after landing as a result of the firefighting activities.
Tests of the two engine-driven generator control panels showed that the
differential control circuitries in each of the panels had detected faults on their
respective a.c. buses, displayed the faults on the control panels, and then tripped each
generator off its respective bus. The differential current circuitry of a generator control
panel is designed so as to trip the associated generator from the line within 0.1 second
after detecting a
to IO-ampere fault current.
During the test of the generator
control panels, the protective trip occurred within the prescribed limits.
The static inverter used to provide emergency a.c. power was tested
.
functionally by
Safety Board at Air Canada’s Maintenance Base, Dorval, Quebec,
Canada.
power terminals were found to be short circuited, and the unit
was torn down for detailed examination and testing.
Two of the eight ‘power transistors
were found to be short circuited. When the shorted transistors were replaced, the
inverter functioned normally.
The Safety Board could not determine the cause of the
.
short circuits.
I
’.
’
~
_
-36-
1.16.3
Flush Motor and Lavatory Components
The flush motor, flush motor components, and various other electrical
components removed from the airplane were taken to Transport Canada’s Safety
Engineering Laboratory in Ottawa,
or to Air Canada’s Maintenance Base, Dorval
International Airport, Quebec, and examined and tested. The tests and examinations were
performed under the supervision of Safety Board personnel.
X-rays of the aft lavatory flush motor were taken before it was disassembled.
The X-rays revealed no evidence of internal melting or shorting of motor components.
Solidified melted metal was observed which appeared to be melted motor housing.
Solidified melted metal deposits were removed from inside the motor, below the rotor.
The motor shaft appeared to be encased in solidified melted metal; however, it could be
rotated very slightly. A wire was still attached to a ground stud inside the motor. When
the stud was removed, the wire broke due to brittleness. There was no evidence’ of
electrical arcing observed on the ground wire or the stud.
A portion of the solidified melted metal which partially encased the
was
.
removed, and the was rotated about The fiberglass insulating material around
the
windings and the silicone glass was intact but scorched. The assembly
was intact and showed no visible signs of electrical arcing.
When the remainder of the solidified melted metal deposit which partially
encased the
was removed, two wire segments were found which had been encased
partially within it. One additional wire segment was loose under the metal deposit. The
source of the metal deposit apparently was a portion of the motor housing which had
melted, flowed downward, and entrapped t-he wires between the motor mounting flange.
When the metal deposit was removed from the mounting flange,
of the
circumference was visible. The fiberglass insulation on the visible portion of the wire
segments was intact, but showed evidence of scorching. Wire splices were found in the
two wire segments encased in the molten metal as well as in the loose wire segment.
The
splices were located just outside the solidified melted metal deposit.
No evidence of
electrical arcing was observed on the exposed wire segments or on the splices. X-rays
were taken of the solidified melted metal deposit and wire segments; the x-rays revealed
no evidence of electrical arcing.
The motor was removed from its mounting flange. The gasket between the
motor mount and the mounting flange was brittle and scorched. The remaining solidified
melted metal from the motor housing appeared to have flowed down around the rotor
shaft and formed a deposit around the motor mount. The mating face of the motor mount
was intact and showed no evidence of melting.
The motor’s
assembly was removed and inspected. The cable clamp that
routes the power leads from the timer harness to the motor was still intact around
four wires. The mounting hardware was still attached to the clamp; however, the
attachment point at the motor housing could not be found; it apparently had melted away.
The four power leads still were routed to the
windings. X-rays of the
assembly revealed no evidence of electrical arcing or internal melting.
The
windings, where the power leads were connected, was partially disassembled. The enamel
insulation around the wire used in the
windings had been melted away, and bare
copper wires were exposed.
No evidence of arcing was observed at the power lead
connections or in the windings.
.
. .
Six segments of wire from the flush motor timer harness were found in the
pump well next to the motor mount.
Four of the segments had splices in them; the other
two wires had no splices.
The two wire segments that had not been spliced were
considerably shorter than the four segments with splices.
The insulation sleeving around
the splices appeared to have been melted away; however,
the mechanical connections of
the splices were intact. The splices showed no evidence of electrical arcing or shorting.
However, three of the wire segments showed signs of electrical arcing-the wire ends
were melted into the shape of a globule.
The metallic globules, which were located on the
wire segments at the point where the wires traversed the lightening hole in the partition
between the toilet and amenities section, were examined using a scanning electron
microscope
The X-ray energy dispersive analysis indicated that they were copper.
The flush motor wiring harness was examined closely for any evidence of electrical arcing
or shorting.
The connector at the flush timer end of the harness showed evidence of
damage from intense heat.
Except for a small tear 25 inches from the connector, the
sleeving which surrounded the wires was found intact from the connector to a point 35
inches away. The teflon insulation on the individual wires within the harness was found
intact from the connector to a point 38 inches away.
From that point to the end of the
harness, the sleeving and teflon insulation had melted away progressively, and toward the
end of the harness, bare wires were exposed.
No indication of electrical arcing or
shorting was observed on any of the wires.
The
phase-A, phase-B, and phase-C flush motor circuit breakers
were removed from the cockpit and were X-rayed, revealing no internal damage. During
a functional test, all three circuit breakers tripped when the electrical load exceeded the
S-ampere rating. The circuit breakers were connected to a power source and load and
subjected to a
lo-ampere current overload.
Only the phase-A circuit breaker
exceeded the time limit designated in the specifications. before it tripped. The Safety
Board could not determine the reason for the failure of the phase-A circuit breaker to
meet its specifications; however, all three circuit breakers showed evidence of damage
due to an external heat source.
.
The flush motor timer was examined.
A continuity check of the timer’s
three-phase power relay contacts showed that they were open -- the normal position of
the relay when the flush button is not engaged.
A 3-foot, S-inch portion of the flush timer’s wiring harness, from the
connector at the timer to just outside and forward of the lightening hole in the vanity
structure, was removed and examined for any evidence of electrical arcing and shorting.
The first 2 feet of the sample was relatively intact.
From a point 2 feet 6 inches to a
point 3 feet 5 inches from the connector, the outer insulation layer of the individual wires
had been gradually melted away; however, the fiberglass inner insulation remained intact.
The wires were bare of insulation over the last 5 inches of the harness, and when the
harness was removed, the wires broke due to brittleness.
No evidence of arcing or short
circuiting was observed on any of the wires.
The following electrical components were removed from the lavatory and
examined for electrical arcing and short circuiting:
the lower mirror light assembly, the
upper mirror light and dimming switch assembly, the aft reading light transformer and
circuit breakers, the razor outlet converter, the aft attendant panel, and the aft
miscellaneous lights control transformer. All of these components and-their associated
wiring were damaged by heat, and portions of some of them were missing; however, no
evidence of electrical arcing was observed on any of the wires.
.
’F.
’
-3a-
In addition, several samples of spliced electrical wiring were removed for
examination from the forward side of the aft pressure bulkhead.
of these samples
exhibited evidence of exposure to extremely high temperatures. The insulation covering
was missing from all of the splices; however, no evidence of electrical arcing or shorting
was observed.
1.16.4
Motor Seizure Test
At Air Canada’s Dorval Maintenance Base, the Safety Board simulated the
conditions produced by a seized or frozen flush motor assembly drive shaft.
A Western
Gear Motor, Model
identical to that on Flight 797, was connected to a test
fixture which provided 115-volt a.c. three-phase power through 5-ampere circuit breakers
connected to the motor power leads. The rotor shaft was locked, the motor was operated,
and the internal motor temperature and motor case temperature were measured.
The internal motor temperature began to rise as soon as power was applied.
At 1 minute 30 seconds after power was applied, smoke was visibly emanating from
around the motor cover plate.
At this time, the motor temperature was
At 6
minutes 15 seconds after power was’ applied, maximum rotor temperature of
F was
reached.
A few seconds later, two phases of the motor
windings opened. At 7
minutes after power application, maximum motor case temperature,
F, was reached,
at which time both rotor and case temperatures began to decrease. Both temperatures
continued to decrease until the test was terminated. At 27 minutes 16 seconds after
power application, the remaining motor
winding opened. Since no further current
flow was observed in any of the three-phase motor leads, the test was terminated. The
rotor temperature observed at this point was
F and the case temperature was F.
The circuit breakers did not trip test. The maximum current flow
recorded during the test was 1.85 amperes’ per phase.
After the motor assembly cooled, it was examined.
revealed that
the rotor was heavily darkened around the circumference of the rotor area which aligns
with the
The rotor was intact and appeared undamaged. The exhibited a
heavily darkened area around its circumference where it aligned with the rotor. X-rays
of the
did not reveal any internal electrical arcing or melting.
1.16.5
Fire and Heat Tests
During the investigation, the Safety Board conducted flammability tests on the
materials contained in the Heath Tecna cabin interior assembly kit. In addition, tests
were conducted to determine the effects of fire and/or heat flux on DC-g-32 lavatory
components, flush motor pump components, wiring bundles, wire insulation, and waste
materials.
These tests were conducted at the FAA Technical Center, Pomona, New
Jersey.
Cabin Materials Burn
The
materials contained in the Heath Tecna Kit
were subjected to the current standard Bunsen Burner tests as set forth in 14 CFR 25.853.
(See appendix
All of the materials tested met prescribed standards.
A piece of polyurethane seat cushion, similar both in time of service and in
composition to the seat cushions on Flight 797, was subjected to vertical and horizontal
Bunsen Burner tests. The material failed the vertical test but passed the horizontal test.
‘.
-
’s
-39-
.
l
. .
.
.
.
The FAA project manager in charge of full-scale fire testing at the Technical Center was
asked why the material had failed the test after only 18 months in service. He speculated
that the particular piece of foam tested had lost some of its fire-retardant capabilities
because of the effects of wear and body moisture on the outer surface. He testified at
the Board’s public hearing that Center technicians had encountered similar failures when
testing older seat cushion materials and that he did not believe that the degraded
capability of the seat cushion would have contributed to the propagation of the fire on
Flight 797.
Cold Air Supply Line Tests. --The susceptibility to heat of a cold air supply
line similar to the one that had melted through on Flight 797 was evaluated. A sample
cold air supply line was placed near an
y powered heat element capable of
producing a radiant heat flux of about 7 BTU/ft -sec. The amount of heat flux was
controlled by placing the subject line at predetermined distances from the heat source.
The cold air supply line was capped at one end, and as the line heated, a constant internal
pressure of 1 psi was maintained by manually opening a relief valve. During each test, the
aluminum cold air supply line was held stationary for about 15 minutes until there was no
noticeable increase of pressure within the line.
Three tests
conducted. The heat fluxes in the cold air supply line were
about 2, 5, and 7 BTU/ft -sec.
There was no evidence of heat damage to the line and
hardness test results showed that the line remained within its specified tolerances.
The
generated from burning paper towels were also evaluated. Three
paper towels were crumpled by hand and ignited by a match, and
heat flux was
measured.
The maximum measured heat flux was about 4 BTU/ft and the
temperature was about
F. The heat generated by burning towels and without
airflow was also evaluated; the maximum heat flux remained at 4 BTU/ft -sec.
DC-g-32 Lavatory Mockup Tests
.--A partial DC-g-32 lavatory was
constructed with actual airplane hardware.
Tests were conducted to explore the effects
of radiant heat from a toilet flush motor on adjacent lavatory components, and the
effects of fire impinging on the flush system power harness. Thermocouples were placed
at various locations within the mockup to monitor temperatures. No air flow was used for
these tests.
A Western Gear Flush Motor, Model
Serial No. 2984, was used in the
first test.
In order to simulate an overheated motor condition, the rotor shaft was
mechanically restricted.
The motor was then mounted on a pump assembly utilizing a
cast aluminum mounting flange and installed in the waste tank. Except for the restriction
of the rotor shaft rotation, these conditions simulated exactly the assembly of the
accident airplane.
Three-phase
a.c.,
400 Hz power was supplied to the motor through a
wiring harness of the same length and type as that installed in the accident airplane. The
power supply end of the harness was protected by the use of three
circuit
breakers. The wiring harness in the lavatory mockup was routed similarly to the
installation in the accident airplane. A flush timer was not used for this test; however,
the power harness was routed to the location, in the vanity, where the timer was installed.
The power harness was mated to the flush motor through the use of a junction box and
normally used connectors.
’
’F.
’F
’F.
Three-phase power was applied to the motor until all three windings of
the motor had failed open. Light smoke was observed coming from the toilet bowl about 5
.
minutes 45 seconds into the test. Subsequently, the phase-A, -B, and -C windings
failed open at 6 minutes 40 seconds, 17 minutes 45 seconds, and 11 minutes 25 seconds,
respectively.
The maximum motor case temperature --
F -- was reached at 9
minutes 45 seconds after power was applied.
The maximum temperature on an adjacent
lavatory component (toilet shroud) of
F was reached 13 minutes 55 seconds after
power was applied. Twenty-five minutes of data was recorded at which time the test was
terminated. No evidence of any deformation, discoloration, or overheating of any of the
vanity or waste tank components was observed.
The flush motor was removed from the lavatory mockup and examined. Oily
residue was found below the motor case on the mounting flange. When the top cover plate
of the motor was removed, evidence of overheating of the cover gasket, rotor, and
assemblies was apparent.
The rotor appeared locked when hand rotation was attempted;
but when additional torque was applied to the shaft, the rotor turned freely. The rotor
assembly was removed and evidence of arcing at numerous points along its top outboard
face was observed. The
assembly was removed from the motor case and examined.
The lower gasket under the assembly was intact; however, oily residue was
observed. An electrical continuity check was made of the windings; no continuity
was observed phase-to-phase or each phase-to-neutral.
However, there was high
resistance continuity between the phase-B winding to the
assembly case, and the
phase-C winding to the case. These resistance readings were greater than 2 megohms and
1.3 megohms, respectively. The wire harness which provides power to the assembly
showed no sign of damage.
the second test, a flush motor housing containing a controllable electric
heating element was placed in the waste tank in the lavatory mockup. The voltage to the
heating element was then increased until the temperature on the outside the motor case
reached a maximum of slightly above
This temperature exceeds by nearly 100
percent the highest temperature that has been reported on this type of motor.
The maximum temperature of
was reached 42 minutes into the test.
Forty-four minutes into the test, the maximum temperature on an adjacent lavatory
component
-- the toilet shroud
-- had reached
Power was removed from the
heating element.
The only observable change to any lavatory component was in a
area in the aft outboard corner of the toilet shroud directly above the flush motor. This
area was deformed upward to a height 2
inches above the shroud support bracket on
the forward face of the vanity.
There was no discoloration, melting, or any other
deformation of the shroud; however, some adhesive material, used to bond a doubler to
the underside of the shroud around the toilet bowl cutout, flowed from the bond line and
dripped onto the flush motor case.
The adhesive also dripped onto the elastomeric hose
running from the flush pump mounting flange to the toilet bowl.
The motor case remained intact; however, obvious signs of overheating were
present. The name plate on the top cover was completely blackened, and the inside of the
cover was brown around the outer edge, becoming lighter toward the center.
Flush Motor Power Harness Fire Test
.--A test fixture was constructed using a
piece of
OSO-inch
sheet aluminum 21 inches by 6 inches. A 1 diameter
lightening hole was cut in the fixture, an ‘alligator grommet was installed. around the
‘is
_
The first chafe test was conducted with the nylon alligator grommet around
the surface of the lightening hole.
After 10 minutes of rubbing, the outer heat shrink
.
cover of the harness was penetrated.
The outer nylon insulation of two of the eight wires
.
in the harness had been chafed slightly; however, the insulation had not been penetrated.
The same test was conducted with the alligator grommet removed. After 2
minutes of rubbing, the heat shrink outer covering had been penetrated and the nylon
outer insulation of one wire was chafed. After 3 minutes, the insulation of one wire was
penetrated exposing bare conductors; however, no electrical activity was observed. After
4 minutes, the exposed wire broke.
After about 8 minutes, electrical arcing occurred
between another wire in the harness and the structure of the lightening hole, but none of
the circuit breakers tripped. The test was resumed, and 2 seconds later, the exposed wire
severed at the point of contact with the lightening hole. The phase-B circuit breaker
tripped simultaneously with the severing of the wire.
The method used to expose the wires in the harness during the two tests was
not intended to duplicate what would occur during actual operating conditions had either
the harness support failed, or had the alligator grommet been missing, or both. Given the
light weight of the harness, even had the support and the grommet been missing, the
harness would not have been subjected to the abuse during actual operating conditions
that it was subjected to during the tests. The tests were conducted to determine what
would occur if the wires in the harness were exposed and to determine the effort required
to abrade the insulation and chafe the wires.
1.16.6
Airplane Cabin Fire Research
Small scale individual fire tests of cabin materials do not replicate the
dynamic range of conditions present in actual cabin fires. Consequently, 1978, the
FAA instituted a research program at its Technical Center involving full scale cabin fire
tests. The tests are conducted in a full scale, wide-body test platform, constructed from
a surplus McDonnell Douglas C-133 airplane.
The purpose of the tests has been to
understand and demonstrate the behavior of cabin materials in a postcrash fire. The
results of these tests were described at the Safety Board’s public hearing by the FAA
project manager in charge of full-scale
testing and are also contained in the Advisory
Group for Aerospace Research and Development Report LS-123, “Aircraft Fire
Safety.”
Of particular relevance to this accident are those data relating to
flashover,
flashfire, seat-blocking, and cabin hazards created by burning interior materials.
The FAA project manager testified that two main phenomena occur when
large fires are expanding; one is the flashover, the other is flashfire, and most often, they
occur in combination.
Flashover in an airplane cabin environment occurs when enough
heat has built up along the ceiling so that the radiant flux down to the materials below the
heat layer reaches a level that is high enough to cause an almost instantaneous ignition of
the material. FAA research has indicated that flashover produces nonsurvivable
conditions throughout the cabin within a matter of seconds.
Flashfire is the burning of combustible gases. According to the FAA
manager, it really a mild
Flashfire occurs when materials in a localized
area burn and emit combustible gases.
The combustible gases, a result of incomplete
combustion, accumulate until they reach a flammable limit and will, if there is a source
of ignition, ignite.
The resultant fire will propogate rapidly, usually at the ceiling where
the combustible gases have collected. With regard to the ignition source, the FAA project
If
‘*
an
”
no
”
Individual units are protected by individual trip-free circuit breakers. These,
when tripped, cannot be electrically reset until cooled. Those units requiring three-phase
supply will have. individual breakers in each phase.
Electrical System Abnormalities
.-The procedures to be followed in the event
of circuit breaker
are contained in the abnormal operation section of the AOM.
Unless directed otherwise in a specific abnormal operation, the pilots may attempt to
reset any tripped circuit breaker. The procedure notes that it may be necessary to allow
about 3 minutes cooling time before a circuit breaker will reset. It also states, the
breaker will not latch or trips immediately after reset, leave the breaker open The
procedure also contains the following note,
circuit breakers protecting a single phase
are trip free.
Manually holding in a breaker which will not latch, will not complete a
circuit.”
The Air Canada Manager. of Flying Operations testified that if a circuit
breaker cannot be reset, flightcrews are instructed to wait for
appropriate cooling
and then try to reset it.
If the reset fails, the circuit breaker is left out. He
testified that
more investigation is required because the electrical power to the
malfunctioning circuit has been cut off and you don’t want to do anything that might
restore
The emergency procedures section of the manual contains a checklist for
detecting the source of electrical smoke or fire. The 4-page checklist essentially requires
the pilots to shut down each of the airplane’s electrical systems, assess the quantity. of
smoke, and then turn on each component of the system one at a time in order to ascertain
which component is the source of the smoke.
Smoke Removal Procedures.--The following abnormal procedures concern the
removal of smoke from either the cockpit or the passenger cabin of the airplane:
The procedure used by Air Canada for removing smoke from the passenger
cabin by opening the right forward galley door and aft pressure bulkhead door was
developed by McDonnell Douglas.
Its DC-9 cabin smoke removal test flights showed that
when the right forward galley service door and the aft pressure bulkhead door were
unseated, smoke introduced into the cabin by smoke generators was forced forward and
out of the galley service door.
This flow pattern was the result of the differential in
airflow outside the two doors. The higher local airflow outside the galley service door
produced a lower outside ambient pressure at the galley service door relative to the
ambient pressure inside the
at the aft pressure bulkhead door, thus forcing the
smoke forward and out the galley service door.
The procedure was presented to the FAA
for approval as an emergency procedure and was rejected. According to the FAA DC-9
project manager test pilot, FAA disapproval was not based on the efficacy of the smoke
removal capability, but on the fact that it required a flight crewmember to leave the
cockpit to operate the doors during a period wherein it believed his presence was required
in the cockpit.
According to the test pilot, the FAA flight test personnel were not
authorized to judge whether or not the door operation could be performed by flight
attendants; therefore, based on the foregoing, the procedure was disapproved. However,
the regulations did not preclude the manufacturer from providing the procedure to DC-9
operators as a manufacturer’s recommended procedure.
McDonnell Douglas provided the
procedure to its DC-9 operators, and Air Canada, with the approval of
Canada,
incorporated the procedure in its AOM.
”
there’s
_
_
’
-
-45-
.
During the public hearing, a fire protection engineer who had participated in
,
the on-scene investigation as a member of the Safety Board’s structures group testified
that in his opinion, if these doors had been opened as envisioned in the abnormal
procedure,
a very strong potential that (the forward airflow) would have pulled
the fire out of the lavatory into the cabin and certainly would have moved the smoke
forward and faster over the passengers heads.”
He stated that it would have endangered
the passengers and also the safety of the airplane.
According to the captain, and the Air Canada manager of flight operations,
the company did not advocate using the cabin smoke removal procedure unless the fire
was out.
The captain also testified that the procedure required the first officer to leave
the cockpit in order to either supervise the opening of the doors or to open them. In the
circumstances of the accident flight, given the airplane’s electrical and mechanical
problems, he believed the first officer was needed. in the cockpit. Therefore, the captain
did not consider ordering the crew to use the cabin smoke removal procedure.
Emergency Descent Procedures
EMERGENCY DESCENT
rtructurxl integrity
in doubt, limit
aond (II
and avoid tiigh m~nouvorlng
l
l
. . . . . . . .
. .
. . . . . .
. . .
l
AUTO PILOT
. .
. . . . . . . . . . . . . . .
. . . .
. . . . . . . . . . . .
. . . .
. .
. . . . . . . . . . . . . . . .
.
SPEED BRAKES
. . . .
. .
. . . . . . .
. . . .
. .
. .
. . . . .
. .
. .
. . . . . . . . . . . .
0.70 M/320-350
Do not l
Do not
rotrim horlrontxl8tablllzor.
When at Flight Attondantr
nstrictad operrtionr
thorrpy aa
Cebin altitude
below lO.DOD
. . . . .
. . .
CREW OXYGEN
MASKS STOW
. . . .
. . .
. . . . . . . . .
. .
. .
Air Conditioning and Pressurization
Airplane pressurization and air
conditioning is provided by the left and right air conditioning and pressurization packs
(packs)
are supplied bleed air from their respective engines. Normally, the right
system supplies the air requirements of the cabin, and the left system supplies the air
requirements of cockpit.
The AOM states,
distribution of conditioned air to the flight
compartment is designed as a continuous smoke removal system.”
With the packs
operating, conditioned air enters the, cabin through ceiling outlets. The air is exhausted
through perforated panels at the cabin floor line and through the left and right tunnels in
the cargo bays to the outflow valve. The outflow valve controls the exhaust rate to the
atmosphere. The airflow in the cockpit is similar to that of the cabin. Opening the
.
outflow valve, either manually or electrically, causes the airplane to
Engine bleed air is furnished to the packs from the engine’s 8th stage (low
pressure) or 13th stage (high pressure) bleed air manifolds, depending upon engine power
settings and the demands on the pneumatic supply.
If the pack supply switch is in the
.
auto,”
”
heard
position and the low pressure bleed air decreases to about 18.5 psi or less, the
system’s augmentation valve opens and high pressure bleed air is introduced into the
system to maintain 18.5 psi,
When bleed air pressure returns to
psi, the
augmentation valve will close. (The augmentation valves are powered by the left and
right
buses and will fail closed if electric power is lost. The other two valves in the
system
-- the flow control and pressure regulation valves -- are powered by the d.c.
emergency bus and will fail open if electric power is lost.) During a descent, if the
engines are at flight idle thrust, high pressure bleed air is required to maintain the
minimum system pressure. At level flight thrust settings, at any altitude, the low
pressure bleed air will maintain the pneumatic pressure at or above 18.5 psi.
At 33,000 feet, in level flight and with the pack switches in
the entire
cockpit and cabin air volume (4,391 cubic feet) is exchanged every 2.2 minutes. During
descent to 3,000 feet with the engines at flight idle thrust, the time required to exchange
completely the cabin and cockpit air varies from 2.3 to 2.7 minutes. In level flight at
3,000 feet, the exchange would
completed within 2.2 minutes.
The positions of the flow control and pressure regulation valves during flight
could not be determined from their positions after landing. Regardless of the position of
their respective pack supply switches, with no pneumatic pressure in the systems and with
all electric power off the airplane, these valves would have returned to the open position.
(There was no evidence that smoke entered the cabin through the air conditioning system.)
With regard to the aft lavatory, the air supply enters the aft lavatory through
the louvered door. and from the overhead duct. It is vented out of the lavatory through
the floor and the. aluminum vent tube within the toilet shroud. In addition, the flow of air
in the amenities section of the vanity flows forward into the toilet area below the shroud
and is also vented out of the airplane through the vent tube. Given this flow pattern, the
FAA project manager was asked what effect opening the lavatory door and then chopping
away the lavatory wall panels would have had on the fire.
He testified that since he did
not know the exact location of the fire or its intensity, it would be difficult to determine
what effect opening the door would have had on the fire. **However, because there is a
vast amount of air supplied through the vents in the lavatory,‘? he did not believe that
opening the door would have provided. a large amount of additional
to the fire.
Consequently, opening the door might have allowed some smoke and some flames, if there
were flames in the area, to escape for a short period of time while the door was open.
With regard to removing the lavatory paneling, he testified that the lavatory area is not a
closed container; there is already airflow in the area behind the sidewalls.
He suspected
that opening the area by removing the paneling “would not sufficiently intensify (the fire).
It may initially get it burning a little bit, but anything that you could do to get at the fire
to fight it, if you had the means of doing so, and you could do it, should be
Aft Lavatory
Passenger interviews indicated that the left aft lavatory was used several
times during the flight. The last known passenger to use the lavatory stated that she had
operated the toilet while she was in the lavatory and that it had operated normally, and
she heard no unusual noises while the toilet was operating. She observed nothing unusual
,
while she was in the lavatory and when she exited there was a male passenger waiting to
enter. The male passenger could not be identified.
According to the passenger, about 25 minutes after she left the lavatory, she
a commotion in the back of the- airplane.”
The passenger also stated that she does
not smoke.
’s
If
”
“team
-
‘..
block displays the airplane’s call sign, ground speed, and last assigned altitude. The ARTS
III computer in service at the Cincinnati TRACON could not track, display, or attach a
data block to a primary target; however, the airport surveillance radar was capable of
displaying an airplane’s primary target. Since Flight
transponder was inoperative, an
automated handoff of the flight from Indianapolis Center to the Cincinnati TRACON was
not possible.
Therefore, the controllers used the interfacility
to hand off the
flight manually.
FAA Order
7110.65C, “Air
Traffic
Control,
January 21, 1983, contains the
recommended procedures for transferring radar identification of an airplane from one
controller to another or from one facility to another.
Controllers are required to be
“familiar with the provisions that pertain to their operational responsibilities and to
exercise their best judgment if they encounter situations that are not covered by it (the
. .
Order).”
The procedures require the controller initiating a manual handoff to convey to
the receiving controller that he has a “handoff.”
Thereafter, he shall provide the
receiving controller with the position of the target relative to a fix, a map symbol, or a
known radar target which is displayed on the screens of both the receiving controller and
the transferring controller. The controller initiating the handoff should provide the
. airplane identification, its assigned altitude, any restrictions, and if applicable, whether
the airplane is climbing or descending. He should also advise the receiving controller of
pertinent information not contained in the data block or flight progress strip. Pertinent
information includes assigned heading, airspeed and altitude restrictions, observed track,
and beacon code if different from that normally used or previously coordinated. The
receiving controller shall, in turn, insure that the target’s position corresponds with that
given by the transferring controller or that there is an appropriate association between an
automated data block and the target being transferred before accepting a handoff.
Paragraph 661b of the Order states,
identification is questionable for any
reason, take immediate action to reidentify the aircraft or terminate radar service.”
With regard to identifying a primary target, paragraph 654 states, in part, that a primary
target can be identified by “observing a target an identifying turn or turns of
30 degrees or more, provided
.
. one aircraft is observed making these
Three methods of identifying beacon targets are provided in paragraph 655.
The controller can either request the airplane to activate the identification
function of the transponder and then observe the identification display; request the
airplane to change to a specific discrete or nondiscrete transponder code and then observe
the display change; or request the airplane to change its transponder to “standby” and
observe the loss of the beacon target, then request the airplane to return the transponder
to normal, and observe the reappearance of the beacon target.
The Louisville high controller said that the handoff of Flight 797 to the
Cincinnati
was a
effort,
between him and the LEX-D controller. The
LEX-D controller’s position, which was located across the aisle from the Louisville high
controller’s position, is a
position and is responsible for coordination between
sectors, flight plan updating, and computer inputs.
The LEX-D controller, who was
radar-rated, had not been in direct communication with the Louisville high controller;
however, he had overheard him discuss the emergency with Flight ‘797. The LEX-D
controller did not cross the aisle either to observe the Louisville high controller’s
-
”.
’s
radarscope or to talk with him.
He said that he tried to configure the radarscope next to
his position to obtain a transponder code and data block on Flight 797, but his initial
. .
attempts failed. He knew the flight was being vectored to land at Cincinnati; therefore,
at
he called the Cincinnati TRACON on the to alert the Cincinnati
approach controller of an impending handoff in his southwest sector.
He also asked his
radar controller seated at the radarscope next to his position to program the computer to
display Flight
data block so that he could get its transponder code for the handoff.
The LEX-D controller stated that he also had overheard Flight
receive
clearance to descend to 5,000 feet and to turn to He knew the airplane had an
fire, but he was not aware of its electrical problems. According to the Center’s
standard operating procedures, Flight 797 should have been handed from Louisville High
Radar Sector to the Evansville/Nabb Low Radar Sector and then to Cincinnati approach
control. The LEX-D controller said that he independently’ made the decision to effect a
direct handoff of Flight 797 from the Louisville high controller to the Cincinnati approach
controller.
The LEX-D controller said that, when the approach controller accepted the
handoff, he did not hear the approach controller say the 0662 transponder code; he only
heard him say the altitude. He said that, at that moment, he was talking directly to his
radar controller seated beside him and he believed that the
in his response was
directed to his radar controller. He was not aware that he had also transmitted the word
.
to the Cincinnati approach controller when he called later to correct the altitude.
According to the LEX-D controller, by
Flight primary target
symbol and data block were being displayed on the adjacent radarscope. However, he did
not tell the approach controller that Flight 797’s transponder was inoperative because he
was not sure that it was, in fact, inoperative.
He stated that a beacon can be missed for
several sweeps of the radar antenna,
and in the interim, until it is reacquired, the
will display a primary target symbol on the radarscope.
The LEX-D controller stated that there were no beacon targets near Flight
797 during the handoff. He believed that he had pointed out the correct target and that
the Cincinnati approach controller had accepted the target he had pointed out.
Thereafter, at
the LEX-D controller called the approach controller on the
and told him that Flight 797 had been assigned a heading and that the
approach controller repeated the heading and signed off with his operating initials. At
the LEX-D controller told the approach controller that the flight was descending
to 5,000 feet.
When the Cincinnati approach controller was alerted to the impending handoff,
he saw a westbound beacon target in the southwest sector of his radarscope. The target
was above the
upper altitude filter limit of his scope and was displayed as an
asterisk with no data block. As a result of the
and transmissions from the
LEX-D controller, he knew he was accepting an emergency but he also expected to
receive a transponder coded handoff.
He used his computer trackball
12/ slewed
to the
target he had observed earlier, entered the position into his
and received a
partial data block containing a 0662 transponder code and the airplane’s altitude --
FL 350. (Under these conditions, the partial data block would only be displayed for about
three sweeps of the radar antenna.)
He advised the LEX-D controller. of the code, the
altitude he had observed, and when LEX-D answered “Yeah, thirty-three now,
five southwest,” it confirmed his belief that he had identified and accepted Flight 797.
”
the
’d
kept
”
-
_
The
approach controller said that there were no other beacon targets or
primary targets
near the 0662 code. He said he was aware of the altitude difference, but
he also was aware that Flight 797 had declared an emergency and was probably
descending. Therefore, the mode C altitude data might be lagging. At the time of the
handoff, he had not been told the flight’s assigned heading. Believing he had identified the
correct airplane, he assigned a radio frequency and waited for Flight 797 to contact him.
The approach controller stated that after he had requested the
turn and
saw no response from
target he had been observing, he observed a partial data block
containing an Air Canada 797 identification tag in the vicinity of the secondary target he
had initially identified as Flight 797; the TRACON supervisor also stated that he had seen
this data block. About 1911, the Evansville/Nabb D controller had
an amended
flight plan for Flight 797 into the Cincinnati ARTS
equipment. The flight plan, as
amended, contained Flight
assigned transponder code, changed its destination to
Cincinnati, and stated that the fli ht was descending to 10,000 feet. Had Flight 797’s
transponder been operating, the
4
would have configured the Cincinnati ARTS III
computer to accept an automated handoff at the geographic point where the computer
was programmed to accept handoffs from the
sector.
However, since
Flight
transponder was inoperative, the ARTS III computer could not locate a beacon
target to associate with the
data block and the merely placed the data
block on the approach controller’s radarscope. The data block, because it did not match a
properly coded beacon target, was displayed only for about three radar sweeps.
The
a controller was asked, based on his emergency procedures
training, how far
rom the threshold of runway 36 he would have had to have placed Flight
797 in order for it land on that runway.
He testified that it would depend on the weather.
the descent was conducted in visual flight conditions, he thought that pilot might
be able to descend from five or six thousand feet from a point ten miles south of the
airport at a slower speed and complete a landing.
If it were an IFR (instrument flight
rules) approach,
want to be level at twenty-five hundred feet or three thousand feet
maybe seven or eight or ten miles from the airport.
In this situation, it was difficult for
me to judge how quickly the aircraft could descend (and) how tightly he could turn.” The
controller testified that he wanted to avoid vectoring Flight 797 to runway 36, have it
arrive too high and too close to the runway threshold to complete the landing, and then
have to circle the airport in order to land on another runway.
The controller was not familiar with nor was he required to be familiar with
either the indicated airspeeds or the descent rate capabilities of a DC-g-30 airplane
during an emergency descent.
He also said that even if he had identified Flight
primary target, earlier he would not have turned the flight away from the airport and
toward the south in order to space it to land on runway 36. He would have
him
going for the airport at all
1.17.6
ATC Radar Data
Data Analysis Reduction Tool (DART) radar data information was obtained
from the Indianapolis ARTCC.
The data included airplane position information and
available mode C altitude information for Flight 797 and for Continental Airlines Flight
383, which’was transmitting on code 0662. The Safety Board’s laboratory reconstructed
the ground radar tracks of both Flights 797 and 383.
Runways
and
A movable position identification device available to the controller to identify radar
targets on his radarscope.
I
’C),
_’
.
.
18-36 the Greater Cincinnati Airport were digitized and plotted with the ground tracks.
Times of key transmissions and selected key events were included along Flight 797’s radar
ground track. The last radar fix retrieved from the DART’s data occurred at
and
the last transponder information from Flight 797 was received at
(See figure
.
1.17.7
DC-9-32 Descent Performance
According to the manufacturer,
assuming a descent at flight idle thrust,
landing gear retracted, flaps/slats retracted, speed brakes extended, at a temperature
warmer than international standard atmosphere temperature (ISA a final
descent weight at 3,000 feet of 81,600 pounds, and at airspeeds of Mach and 310 KIAS,
Flight 797 was capable of achieving the following descent rates:
at 30,000 feet
7,800 fpm
at 20,000 feet
5,700 fpm
at 5,000 feet
5,100 fpm
The time required to descend from 33,000 feet to feet was 5 minutes 11 seconds and
the still air distance was 34 nmi.
Since the touchdown zone elevation of runway 27L at the Greater Cincinnati
International Airport was 875 feet, additional time would be required to complete the
descent from 3,000 feet, decelerate from descent speed, configure the airplane for
landing, and fly the final approach.
The Safety Board constructed a descent model
containing the time required to complete these phases of the descent and landing. The
model is based on the following assumptions:
Phase of Flight
Time
1.
Descend from 3,000 feet to
2
minutes
2,000 feet at 500 fpm and
decelerate from 310 KIAS to
200 KIAS.
2.
Extend the landing gear, and
1 minut e
extend the flaps incrementally;
stabilize the airspeed at each
flap increment.
3.
Final approach-descend from
1 minute 30 seconds
2,000 feet to 875 feet at
750 fpm.
Based on the times contained in this model and the manufacturer% performance data, the
total time required to descend from 33,000 feet and land on runway 27L was 9 minutes 41
seconds.
The field elevation at Standiford Field, Louisville, Kentucky, is 497 feet. The
Safety Board estimated that 10 minutes 11 seconds would have been required to descend
from 33,000 feet and land at Standiford Field.
_
-
W,tfl
”
’RE
IMOP
‘,RT.
WNOER
~NAL
s~.
‘NO
NOWE
‘
’RE
lSSSr7S
’“OUSAN0.
F
POINT
l *1 ( 01 :41 AC
l
214.0
Figure 13.-Ground radar tracks for Flights 797 and 383.
’
-
I
As stated earlier, the captain had inadvertently extended the speed brakes to
.
the ground position.
According to the manufacturer, during the emergency descent at
Mach and then at 310 KIAS, the airflow over the wings would blow the speed brake panels
down to the position they would have assumed had he extended them to the position
prescribed in the flight manual. Thus, despite the mispositioning of the speed brake
control lever, the airplane’s descent rates during the emergency descent would have been
essentially the same as those cited above.
At
Flight 797 reported that it was leaving FL 330. Thereafter, it
reported it was at 8,000 feet at at 2,500 feet at and, at 2,000 feet at
Based on these altitude the average rates of descent obtained during
the emergency descent were as follows:
between
and 6,410 feet per
minute between and 2,500 fpm; and between and
536
fpm.
2.1
General
The airplane was certificated, equipped, and maintained in accordance with
.
Transport Canada and U.S. FAA regulations and company policies and procedures. The
flightcrew was qualified and certificated properly and the flight attendants were qualified
for the flight. Each flight and cabin crewmember had received the training and off-duty
time prescribed by Canadian regulations. There was no evidence of any preexisting
medical or psychological conditions that might have affected the performance of the
and cabin crews.
-Involved air traffic controllers were certificated properly, and
each controller had received the training and off-duty time prescribed by FAA
regulations. Accordingly, the Safety Board directed its investigation to the ignition and
propagation of the fire; to ATC procedures; to the performance of the pilots and flight
.
attendants after the fire was discovered; and to factors which affected the survivability
of the passengers and crewmembers.
2.2
Ignition.-The evidence substantiates a conclusion that when the smoke was
detected by the
ght
attendants, there was a fire located within the vanity and/or the
toilet shroud in the aft lavatory. Therefore, the Safety Board tried to identify
possible
ignition sources in this area. Given the location of the fire at the time the smoke was’
discovered,
Safety Board identified five possible ignition sources: an incendiary or
explosive device; deliberate ignition; a burning cigarette; the toilet flush motor; or the
flush motor electrical harness. In addition to these five, the arcing damage found on the
feeder cables of the left and right
generators and the maintenance history of the
airplane generating system led the Safety Board to investigate a sixth possible source
of ignition-the generator feeder cables which were routed beneath the floor of the aft
lavatory.
Based on the examination of the physical evidence and the results of the FBI
laboratory analysis, the Safety Board concluded that neither an explosive nor incendiary
device was involved. Also, there was no evidence that the fire was deliberately set.
. .
Since the tests of the materials used in the aft lavatory showed that they met
the fire resistance criteria contained in 14 CFR 25, it would have been virtually
impossible for either a lighted cigarette or sparks produced by electrical arcing to ignite
the materials used in the construction of the lavatory. In order to ignite the lavatory
partitions and walls, some combustible material capable of sustaining high temperature
combustion for the amount time necessary to ignite the lavatory walls had to have burned.
Therefore, in investigating the possibility that the fire was ignited by a burning cigarette,
the Safety Board focused on two areas below the vanity which could have contained
combustible materials and into which a cigarette might have fallen-the sink area
containing the trash chute and receptacle and the adjoining amenities section. Since the
lavatory trash receptacle was the most logical place for combustible material to collect
and since a burning cigarette could fall down the trash chute into the receptacle, the
damage in this area was evaluated.
Had a cigarette started a fire in the receptacle, the
only propagation path out of the receptacle would have been the trash chute. In order for
the fire to reach and short circuit the wires in the flush motor harness, it would have had
to have burned from the top of the trash chute to the lightening hole containing the
harness. The lightening hole where the flush motor harness wires had short circuited was
several feet outboard and
below the top of the chute and the fire damage was not
continuous from the top of the trash chute to the area of the lightening hole. Moreover,
it was unlikely that a fire could have spread from the top of the chute, which had burned
away where it attached to the sink shroud, down to the flush motor harness, shorted the
wires, and then remained undetected for 11 minutes.
. .
. .
Although there. was some evidence of flame damage both in the area of the
trash chute and the receptacle, the automatic fire extinguisher had discharged only
into the chute. The evidence showed that as the trash receptacle was being heated by the
fire, warm air within the receptacle rose into the trash chute, and was trapped in the area
where the heat-activated discharge nozzle was located.
Based on the damage in this
area, the Safety Board concludes that the air reached the melting point of the nozzle
before the temperature below the sink could attain that level and the entire
supply
was discharged into the trash chute.
In addition, since there was no evidence of a
continuous flame path from the top of the trash chute down to the area of intense fire
damage under the vanity, the Safety Board further concludes that the fire did not
originate within the trash chute.
Also the presence of unburned trash within the trash
receptacle further corroborates this conclusion.
The amenities section of the vanity was almost completely destroyed by fire.
The flame patterns showed that the area of most intense burning was in the lower aft
outboard corner of the void space located in the lowest level of the amenities section and
almost directly below the failed cold air supply line.
The fire was so intense that it
melted a part of the aluminum bottom shelf in this area, while the remainder of the shelf
remained intact.
Since the cold air supply line outlet was closed, there was no airflow
through this pipe initially.
However, once the line melted through, airflow began which
supplied air to the fire and caused it to intensify. The cold air supply line break coincided
almost exactly with the most intensely. burned area on the aft bulkhead. Given this
evidence of fire damage and given the fact that the amenities section adjoins the sink
area containing the trash chute and receptacle, the Safety Board attempted to determine
whether a lighted cigarette could have penetrated the void space of the amenities section.
An aluminum partition separated the sink section containing the trash chute
and receptacle from the open space below the amenities section. However, about 4 inches
above the floor there is a
hole in the partition through which loose material
’.
”
,
.
:.flush
‘bus.
-
can pass into the open space beneath the amenities section from the area beneath the
.
sink. Thus, if the fit between the trash chute and waste disposal door or between the
.
chute and trash receptacle had not been secure, it is possible that a lighted cigarette
placed in the chute partition could have fallen from the chute and thereafter rolled or
made its way through the hole into the open area below the amenities section. The fact
that debris can enter this area was illustrated by the discovery of the vial and
maintenance tag below Flight 797’s vanity, and the waste materials found in this area on
another air carrier’s
The Safety Board concludes that the possibility that this
occurred cannot be ruled out.
To investigate the possibility that the fire was ignited as a result of an
overheated flush motor, the Safety Board conducted two tests. The highest temperature
achieved during the tests was
which was not high enough to ignite lavatory
components in the vicinity of the flush- motor.
In addition, the motors used in the tests
were damaged internally by heat. The flush motor involved in the accident did
show
any evidence of internal heat damage or internal failure.
In order for the flush motor to overheat in service, three malfunctions must
take place: the motor must seize; the flush button must be held in or fail in the
depressed, or power on, position; and finally, the timer must either be defective or fail.
(A properly functioning timer will limit the flush motor to a lo-second cycle even if the
button is held in the depressed position.
Once the timer cycles the motor off, the
button must be released and then depressed again to restart the motor.) The
. recovered components of the timer circuitry were tested functionally after the accident
and were found to be operational.
The last known person to use the lavatory did so about 35 to 40 minutes before
the fire was detected. She stated that the flush motor worked properly at that time. She
also stated that another passenger was waiting to enter the lavatory when she left;
therefore, the Safety Board concludes that the flush motor was operating normally within
35 to 40 minutes before the fire was discovered.
Given the facts that
the test
demonstrated that an overheated flush motor would not produce temperatures high enough
to ignite adjacent lavatory materials;
evidence showed that the flush motor had not
failed internally and was not damaged internally by heat; there was no evidence to
indicate that the flush motor timer had failed; and
the flush motor was most probably
operating normally before the fire was discovered, the Safety Board concludes it unlikely
that the flush motor was the source of ignition of the fire.
Another possible source of ignition near the area where the fire was
discovered was the flush motor wiring harness.
The tripping of the three circuit breakers
accompanied by the arcing sounds recorded by the CVR occurred at
The three
circuit breakers tripped almost simultaneously indicating that the circuitry of all three
phases shorted at the same time.
The only evidence of wiring damage was found
the flush motor wiring harness passed through the lightening hole in the partition between
the amenities’section and the toilet section of the lavatory. The damage noted in the
wiring harness at this location could only have been the result of fire and heat, and the
Safety Board concludes that the damage to the wiring which caused the three flush motor
circuit breakers to trip was caused by heat and fire.
Beginning 3 minutes 2 seconds before the three circuit breakers tripped,
numerous arcing sounds were recorded on the CVR.
These sounds were accompanied by
voltage fluctuations and the electrical components which showed these fluctuations, as
well as the flush motor, were all powered by
right Because of the extensive
-
-
fire damage in the area, it could not be determined whether the flush motor electrical
harness had been properly suspended in the lightening hole, or whether the protective
grommet had been installed around the rim of the lightening hole. Based on the available
evidence, a hypothesis that the grommet was missing and that the wires were not
suspended properly before the fire cannot be supported readily by tangible evidence.
However, assuming that this may have occurred, tests were conducted after the accident
which showed that, although considerable force and effort were required to expose the
conductors, the harness could chafe against the exposed edge of the lightening hole. The
tests showed that the chafed wire could arc against the edge of the hole without causing
the circuit breaker to trip. Such an occurrence could account for the unexplained arcing
signals noted on the CVR. A trash fire ignited by this initial arcing could explain how an
external heat source was generated which melted through the insulation of the remaining
wires and caused the circuit breakers to trip simultaneously. Also, tests demonstrated the
simultaneous tripping
the three
circuit breakers’ after a
exposure of the
harness to a external flame. Since the flammable source which fueled
the fire initially could not be identified, the flame used in the test might not have been
representative of the heat level of the initial fire aboard Flight 797. Although the Safety
Board cannot eliminate the flush motor electrical harness as a possible ignition source,
given the facts that (1) the tests demonstrated that a great amount of effort and force
was required to chafe away the flush motor harness and the insulation of the wires in the
harness; and
the fact that the harness would not have been subjected to the abuse
during actual operating conditions that it was subjected to during the tests, the Safety
. .
Board concludes it unlikely that the chafing of the harness wires against the lightening
holes was the source of ignition for a fire in the amenities section of the aft lavatory.
The sixth possibility of an ignition source
a high resistance short circuit on a
generator feeder cable where it passed through the lightening hole in the floor beam
below the lavatory floor at FS 980
was considered primarily because of clear
indications of arcing between the left and right engine generator feeder cables and the
floor beam. These indications consisted of a notch burned into the floor beam at FS 980,
chafed areas on both engine generator feeder cables which exposed the bare conductors at
FS 980, and the physical indications of short circuiting on the exposed conductors of both
feeder cables. Further, this potential ignition source was in proximity to the floor of the
aft lavatory. In addition, this ignition source was given consideration due to repeated
electrical problems with this aircraft prior to the fire and the “electrical arcing” sounds
recorded on the CVR. This arcing was recorded 3 minutes prior to the tripping of the
three circuit breakers on the lavatory flush motor. A decrease in the right a.c. bus
voltage occurred simultaneously as recorded on the DFDR.
For ignition from this source to occur, the cable must have contacted the
aluminum frame at FS 980. This would assume that the conduit was either missing and/or
broken, that the cables were not supported properly at FS 980, and that the differential
current fault would have been of a low order which would elude the protective circuits
designed to trip the generators.
Given these conditions, it appears
that sufficient
electrical energy could be transferred from the cable to heat and eventually ignite the
nylon conduit and propagate a fire to other combustible materials, probably the epoxy
behind the aft and side walls of the lavatory.
The damage and destruction which occurred beneath the lavatory floor in the
vicinity of FS 980 precluded a determination of whether the generator feeder cables were
.
properly suspended within the lightening holes at that location; however, the evidence of
chafing in the fiberglass insulation of the generator feeder cables where they passed
through thelightening hole at FS 980 indicates-that the cables may not have been properly
-
supported, and that they could have sagged and chafed against the floor beam. Further,
although the fault protection circuitry had actuated and had tripped the generators off
the line, this occurred about 14 and 16 minutes after fire damage had caused the three
circuit breakers of the flush motor harness to trip.
Since the protective circuitry only
operates when the current differential between the generator and the bus exceeds about
20 to 40 amperes, it appears that a transfer of up to 40 amperes through the abraded
cable insulation could have provided sufficient heat to have ignited the nylon conduit.
In an effort to examine the possibility of the generator feeder cables as an
ignition source, the electrical system discrepancy reports were reviewed to determine
whether any might have indicated an intermittent electrical short circuit in the generator
feeder cables. The analysis failed to disclose a problem that ‘could be related to such a
condition in the feeder
even though the electrical problems continued up to the
time of the accident flight.
Because of the extensive damage at FS 980, it was not possible to determine if
the electrical wiring was properly secured
protected, nor was it possible to positively
establish a fire pattern which would permit a conclusive determination that the fire
started as a result of a generator feeder cable fault.
However, this possibility could not
be dismissed.
Propagation
.-Regardless of the ignition source, the physical evidence showed
that there was an area of intense burning in the lower aft outboard corner of the lowest
section of the amenities section, and it also showed that the fire propagated forward from
that point. As the fire moved forward from the amenities section, it also burned through
the lavatory walls. allowing the smoke, hot gases, and fumes to rise in the air space
between the lavatory
and the airplane’s outer skin and between the aft pressure
bulkhead and the lavatory’s liner walls. The seams connecting the lavatory side walls and
ceiling walls are not sealed; thus as the smoke rose, it began to enter the lavatory through
the
and ceiling seams, while the fire remained concealed behind the amenities
section and below the toilet shroud.
As the fire moved forward into the area between the toilet shroud and the top
of the waste tank, the hot gases, smoke, and melted plastic were still being vented
overboard through the lavatory vent line.
The vent line exits the lavatory and enters the
tunnel in the aft cargo compartment aft of FS 980.
It is then routed forward below the
generator feeder cable bundles to a point just forward of FS 965 where it exits the
airplane through a venturi in the lavatory access panel.
The hot gases caused the
aluminum vent tube to melt away, thus permitting the entry of
gases into the
floor beam area below the lavatory floor and their impingement on the generator feeder
cable bundles where they passed through the lightening holes in the floor beam located at
FS 980. As a result of the heat, faults developed on the left and right generator feeder
cables, and between
and the protective circuits tripped them off the
line.
The flexible connection for the waste tank flush and fill pipe and its check
valve assembly was located close to the inlet of the lavatory vent line.
When the flexible
connection and the plastic ball in the check valve failed, this stainless steel flush and fill
pipe also became an overboard vent.
The deposits of soot and tar on the access door to
the lavatory service panel and the melted rivets in the pipe connector at the service panel
confirm that the pipe did become an overboard vent.
The path of almost all fires is upward and in the direction of the airflow.
Except for the damaged area below the lavatory floor, the fire damage noted on the
airplane was above the airplane’s floor line.
The lowest and
point of fire damage
was within the amenities section at the outboard part of the lavatory and the forward wall
of the aft pressure bulkhead, respectively.
From this location, based on the damage
pattern, the path of the fire was upward and forward. The direction of airflow within the
amenities section and the toilet shroud was also forward.
The damage below the lavatory was in the tunnel below its floor. While there
was some evidence of fire damage, the damage in this area was for the most part heat
damage, and the lavatory floor had not burned through.
The most severe damage in this
area was located above the failed lavatory vent tube between FS 965 and FS 980. In this
area, the suspension‘and insulation of the generator feeder cables had melted away and
arcing had occurred. Given the proximity of the failed lavatory vent tube to these cables
and the type damage noted on the cables and structure in this area, the weight of the
evidence indicates that the damage in this area was caused by the hot gases from an
existing fire in the lavatory being vented through the tube. This evidence tends to further
corroborate the hypothesis that the arcing of the generator cables was the result of the
fire and not the source of its ignition.
The momentary smoke abatement noted by the first officer and flight
attendant in
and was probably attributable to the dilution
effect of opening the lavatory door and discharging the CO2 into the area, reclosing the
lavatory door, and the almost simultaneous failure of the lavatory vent line and the flush
and fill pipe connections and check valve, all of which increased the ventilation rate
beneath the toilet shroud and accelerated the flow of smoke and gases to the area below
the lavatory floor and overboard.
The rectangular scorched area on the airplane’s outer skin above the left
engine and coinciding with the aft lavatory’s frame channels showed that as the fire
consumed the lavatory structure it used the airspace between the lavatory outboard wall
and the airplane% outer skin as a flue.
The superheated gases progressed up these
channels and forward along the space between ceiling liner and the airplane’s outerskin,
and began to preheat the ceiling panels. Smoke and fumes generated by the fire began to
collect in the ceiling space.
The smoke, fumes, and hot gases then entered the cabin
through the ceiling and sidewall liners and began to collect in the upper portions of the
cabin.
After the captain
Flight 797 on the runway, both forward exit doors
and three
emergency exit windows were opened and an unlimited supply of fresh
oxygen became available to the fire.
With this availability of oxygen, the preheating of
the ceiling panels, and the large quantities of unburned gases in the upper cabin, a back
draft and flashfire occurred, and the fire progressed rapidly through the entire cabin. The
evidence showed that the carpets,
the lower portions of the sidewall panels, and
combustible portions of the lower seat structure, including armrests, did not ignite and
burn, whereas almost all combustible materials above the windowline were destroyed or
heavily damaged by fire including large portions of the airplane structure and skin. The
physical evidence indicated that the fire in the cabin ignited initially near the ceiling and
thereafter the seat surfaces were ignited by the heat radiated from the fire at the ceiling
and luggage rack level. Therefore, the Safety Board concludes that
rather than
flashover occurred.
’
-
4
. .
Based on estimates by firemen and passengers, the flashfire probably occurred
within 60 to 90 seconds after the doors and
exits were opened.
The upper
fuselage skin, after the cabin ceiling’s insulation blanket was consumed by fire and the
cooling effects of the airflow of flight ended, failed rapidly. The absence of significant
smoke stains on the outside of the open exits supports this view, since following the
failure of the upper fuselage skin, the airflow would reverse and go in the windows and
doors and out the ruptured fuselage skin at the top of the airplane.
In summary, although no positive conclusion can be drawn as to the precise
point of origination of the fire, the evidence indicates that the fire propagated through
the lower part of the amenities section of the lavatory vanity. Because of the direction
of the airflow from the areas below the vanity and above the toilet waste tank, the
smoke! fumes, and hot gases were vented overboard and pulled away from the passenger
cabin and the open area of the lavatory, allowing- the fire
burn-undetected for almost
15 minutes. The first noticeable evidence of smoke within the open area of the lavatory
was observed after the fire penetrated the lavatory liner and began to rise behind and
outboard of the liner.
Smoke then began to penetrate the sidewall and ceiling seams of
the lavatory lining as described by the two flight attendants.
2.3
Operational and Survival Factors
The captain and first officer testified that they did not hear the arcing sounds
recorded by the CVR.
The DFDR showed that the arcing sounds were accompanied by
voltage excursions on the right bus.
The airplane’s wiring diagrams showed that the
DFDR and CVR wiring was routed near the generator cable feeder bundles. Spectrum
analysis of the arcing sounds showed that they contained harmonics that were above the
frequency range that would normally be detected by the microphone and preamplifier in
the cockpit area microphone (CAM) channel. Given the proximity of the CVR wiring to
the generator cable bundles,
the Safety Board concludes that the arcing signal was
electromagnetically induced into the CAM’s circuitry and, therefore, not audible to the
flightcrew. Since the arcing sounds which were recorded by the CVR were not heard by
the flightcrew, the tripping of the flush motor’s three circuit breakers at
was the
first abnormal occurrence they noted.
At
the captain tried unsuccessfully to reset the circuit breakers. At
he again tried unsuccessfully to reset the three circuit breakers. Air Canada
flightcrews are taught to make one attempt to reset a tripped circuit breaker. They are
taught that it may be necessary to allow a 3-minute cooling time before a circuit breaker
will
a reset and that circuit breakers protecting a single phase are trip free;
therefore, a circuit cannot be completed by holding in an unlatched single-phase circuit
breaker.
Most important, they are taught that a tripped circuit breaker denotes that the
circuit protected by the circuit breaker is no longer powered. The flightcrew and other
Air Canada flight personnel stated that circuit breaker trips during flight are not an
uncommon occurrence, and the procedure contained in Air Canada’s
allows
personnel to cope adequately with such occurrences.
In this case, the captain attempted to reset each of the tripped circuit
breakers twice; the first attempt occurred almost immediately after they had tripped and
was unsuccessful. He testified that he ‘*thought at the time that the unit (flush motor)
might be overheated so I just continued the routine of the flight. . . and after a certain
time had passed. . .
I attempted to reset the circuit breakers again to make sure. . . . The
circuit breakers would not move.”
Although the captain was unable to detect any
movement of the circuit ‘breakers, the ‘CVR showed that arcing sounds, which were not
if
I
”
-
audible to either the captain or first officer, accompanied each attempt to reset each
circuit breaker indicating that the circuit breaker had moved and momentary electrical
contact had been made. However, once the contact was made, the protective circuitry
caused the breaker to trip again.
Since the fire was already well established, the
attempts to reset the circuit breakers had no effect on the sequence of events. About
11 minutes 26 seconds after the initial trip of the three flush motor circuit
breakers, the No. 2 flight attendant informed the flightcrew that there was a
When the circuit breakers tripped, there was no reason, based on their
training, for either pilot to surmise that an emergency capable of compromising the
safety of the airplane existed.
Although in this instance the failed component could be
checked visually, the abnormal procedures checklist made no distinction between failed
components which are visually accessible to crewmembers and those which are not. The
pilots are required only to allow the circuit time to cool, and thereafter, limit themselves
to one reset attempt. If the circuit breaker cannot be reset, they can assume that the
component is shut down and, if required, perform the action set forth in an applicable
checklist.
While it can be argued that the captain, from the standpoint of passenger
comfort, should have requested a flight attendent to inspect the lavatory and ascertain
the reason, if possible, for the failure of the flush motor circuit, it was a matter of
judgment on his part as
whether he should require this to be done. In retrospect, his
decision not to do so may have forestalled an early opportunity to discover the fire.
Initial actions taken by the cabin crew when the smoke was discovered were
inadequate to assess quickly the origin and scope of the fire. When the flight attendant in
charge opened the lavatory door, he was able to see into the lavatory and observe that the
smoke was emerging through the seams of the aft wall of the lavatory. Even though he
stated that he knew the fire was not in the trash container, he never did open the door of
the sink compartment to inspect visually the trash chute and container. The dispersal of
the
into the lavatory had little or no effect on the fire.
In order for the
extinguishing agent to be effective, it must be applied to the base of the flames.
According to Air Canada procedures, the fire axe should be used,
necessary,” to remove paneling to obtain access to the fire. After assessing the situation
in the lavatory, the flight attendant in charge did not request that the axe be brought to
him from the cockpit because he did not believe he could use it. He testified that he did
not consider using the axe to remove the paneling because,
would have to destroy half
the aircraft to get to
The flight attendant in charge’s testimony also showed that,
while he knew the company procedure, since he had not been shown which paneling could
be removed with the fire axe without endangering critical airplane components, he was
reluctant to use the axe. Moreover, he was afraid that this action might provide a draft
to the fire and accelerate combustion.
Based on the conditions in the lavatory, the Safety Board cannot determine
whether removing the lavatory paneling would have enabled the flight attendant in charge
to expose and attack the fire successfully or whether the removal of the paneling would
have accelerated the propagation of the fire.
Further, given the situation inside the
lavatory, even had the fire axe been brought to the lavatory, the Safety Board is not
.
convinced that the flight attendant in charge could have carried out the firefighting,
activities contained in the Air Canada manual effectively without a full face smoke mask
with self-contained breathing apparatus.
-
1
”
-
II
-61-
After he had dispersed the
CO
into the lavatory and closed the lavatory door,
the flight attendant in charge briefed the officer on what he had done. Also, he told
the first officer that he did not believe the fire was in the trash container but was located
elsewhere.
Neither he nor the first officer told the captain that they had not seen the
fire and that they did not know exactly where it was or how intensely it was burning. On
the other hand, the captain did not question either man about the location or severity of
the fire.
The only crewmembers to observe the conditions in the lavatory were two
flight attendants. The first officer made two trips aft, but did not enter the lavatory. He
retreated the first time because he did not have smoke goggles with him; the second time,
based on the heat he felt on the lavatory door, he decided that it was not advisable to
open the door.
Since the flight attendant in charge and the first officer were not able to
determine the location of the fire, they were not able to assess the severity of the fire.
Consequently, based merely on their assessment of the course of the smoke drifting into
the lavatory though the seams of the lavatory walls, they provided the captain with an
inadequate assessment of the fire’s severity.
The fire was reported to the captain at
and he directed the first
officer to go aft and assess the situation.
About
the first officer returned to the
cockpit and he told the captain,
think we had better go however, he later
testified that at that time he was not thinking of an emergency descent.
Almost
simultaneously, the captain received a series-of optimistic reports from both the flight
attendant in charge and the first officer concerning the smoke conditions in the aft cabin
area, and at the captain directed the first officer to go aft a second time to
reassess the conditions in the aft cabin area.
a result, about 5 minutes 30 seconds
elapsed between the time that the No. 2 flight attendant told the captain there was a fire
in the aft lavatory and his decision to begin the emergency descent. While an actual
flight fire is an extremely rare occurrence, all reports of smoke in the cabin must be
regarded as potentially serious.
However, such reports often turn out to be smoke from
an overheated flushing motor or waste ignited by a discarded cigarette in a trash
receptacle designed to contain a fire,
conditions which are normally identified and
corrected by flight attendants without further consequences. Therefore, the Safety Board
realizes that there is a need to evaluate the situation before deciding on the emergency
action required. However, in this case, the time to make the decision appears excessive
given the circumstances. Most significantly, neither the flight attendant in charge nor
the first officer was able to fix precisely the source of the fire or to assure the captain
that it had been extinguished. The Safety Board believes that a precautionary emergency
descent should have been initiated as soon as it became evident that the fire had not been
visually located and could be attacked directly with extinguishant. This became known at
when the first officer came forward from his first inspection of the aft cabin
area, about 3 minutes before the decision to begin an emergency descent.
At
after the first officer returned from his first trip aft, Flight 797
was about 14 nmi northeast of Standiford Field, Louisville, Kentucky, at FL 330. Had the
emergency been declared at this time and the descent started, performance data indicate
that Flight 797 could have landed at Standiford about
or about 5 minutes 51
seconds earlier than it landed at Cincinnati.
However, given the actual conditions in the
airplane during the descent, it is not realistic to expect the captain to have duplicated the
times and optimum rates contained in the descent performance profile. For example, the
evidence showed that during the actual descent, the total time required to descend from
33,000 feet to 3,000 feet was about 1 minute longer than. that required in the descent
profile. The Safety Board believes that the evidence indicates that it would be reasonable
-
to assume that, given the conditions in the cockpit, the descent and landing would require
1 to 2 minutes longer than the elapsed time shown in the descent profile. Therefore, the
Safety Board believes that, had Flight 797 landed at Standiford Field, the flight most
probably would have landed about 3 to 5 minutes earlier than it did at Cincinnati.
While the research data does not permit the Safety Board to conclude whether
the shortened flight time would have delayed or prevented the flashfire, there can be no
doubt that the decreased exposure time of the passengers to the toxic environment in the
cabin would have enhanced their physical and psychological capability to escape after the
cabin doors and
exit windows were opened.
Consequently, the Safety Board
concludes that the delayed decision to descend contributed to the severity of the
accident.
At
when the captain did declare an emergency, Flight 797 was closer
to Cincinnati than Louisville; therefore, the Louisville high controller vectored it toward
the Greater Cincinnati Airport and cleared it to descend. At
the flight departed
FL 330 and it landed at Cincinnati at In the process, however, there was a faulty
ATC handoff which occurred at
The Safety Board concludes that the faulty interfacility handoff leading to the
misidentification of Flight 797 resulted from the attempt on the part of the controllers at
the Indianapolis Center to expedite the handling of an aircraft experiencing a dangerous
in-flight emergency. The Safety Board believes that the LEX-D controller’s decision to
hand off Flight 797 directly to Cincinnati approach control was a valid exercise of the
controller’s discretionary authority and if handled properly would have eliminated
additional radio frequency changes and decreased flightcrew and controller workloads.
The Safety Board also concludes that by beginning the handoff without direct communica-
tion with the Louisville high sector controller who was handling Flight 797 and by not
including him directly in the handoff procedure, the timely transfer of vital information
between the two facilities was either compromised or never accomplished. The most
important omission was the LEX-D controller’s failure to tell the Cincinnati approach
controller that he was being handed off an airplane with an inoperative transponder.
The Cincinnati approach controller also contributed, though to a lesser degree,
to the faulty handoff. Instead of waiting for the initiating controller to apprise him of the
identification, transponder code, heading, and altitude of the target to be transferred, he
trackballed out to the target he assumed was the subject of the alerting call he had
received from Indianapolis Center at
Thus at when the LEX-D’
controller told him,
got a code for
instead of waiting for the remainder of the
information, he supplied the transponder code to the initiating controller and thereby
contributed to the ensuing communications breakdown.
The primary question presented
by the faulty handoff was whether it delayed the landing of Flight 797. At
when
Flight 797 began the emergency descent, the DARTS data showed that it was about 27
nmi southwest of, and turning toward, the Greater Cincinnati Airport. According to the
optimum performance data, it would require 34 nmi to descend from 33,000 feet to
3,000 feet. In addition, the winds aloft data showed that the winds during the flight’s
descent were from the west at speeds ranging from about 85 knots at FL 330 to about 8
knots at 3,000 feet; thus it would require more than 34 nmi to descend to 3,000 feet. The
evidence was conclusive that regardless of which runway the controller elected to use for
landing, he would have to vector Flight 797 through some type of traffic pattern in order
to land it at Greater Cincinnati Airport.
In this case, he elected to vector it to land on
runway 27L. .
‘.
*
-63-
.
According to optimum descent performance data, it would require 9 minutes
l
41 seconds to descend and land at the Greater Cincinnati Airport. Since the
.
began at
the earliest possible landing time was or 1 minute 23 seconds
earlier than the flight had actually landed. However, since it is unrealistic to expect the
.
captain to have duplicated the optimum descent data during the approach, and given the
small difference between the calculated landing time and the actual landing time, the
Safety Board concludes that the faulty handoff did not delay the landing of Flight 797
appreciably. Since the descent and landing at Greater Cincinnati Airport was accom-
plished expeditiously and since ATC procedures did not delay the landing, the Safety
Board concludes that the only factor which significantly delayed Flight
landing was
the flightcrew’s delayed decision to descend.
During the descent, the smoke in the passenger cabin continued to accumulate;
it was heaviest at the ceiling and became increasingly thick from the ceiling down to the
floor. As the airplane descended, the smoke moved forward in the cabin., In addition,
with the loss of electrical power, the augmentation valve in the pressurization system
closed, and high pressure bleed air was not available.
Therefore, when the engines were
retarded to flight idle for the descent, little or no fresh air was being introduced into the
cabin. Some of the passengers noted that air stopped flowing out of the eyebrow vents
during the descent, confirming that the augmentation valves had closed.
However, when
the airplane was leveled off after the descent and engine thrust applied, the exchange of
air would have been restored, provided that the air conditioning and pressurization packs
had not been turned off.
. .
The captain’s difficulties during the descent were compounded by the condition
of his instruments.
About 8,000 feet during the descent, the emergency inverter
was lost. The Safety Board could not determine the cause of the failure; however, with
the loss of the inverter, the emergency
bus was lost.
The airplane’s
horizontal
. .
situation indicators, and radio magnetic indicators became inoperative, and the only
attitude indicating instrument available was the small emergency standby ADZ. The
captain flew the latter part of the descent, the traffic pattern, and the landing using the
standby AD1 and his airspeed indicator.
In addition, with smoke ‘entering the cockpit, he
had difficulty seeing the instruments.
The damaged cockpit door was not closed after the first officer reentered the
cockpit. The captain had not ordered the door to be left open and was, in fact, not aware
that it was open.
The first officer had decided to leave the door open because it
facilitated communication with the cabin crew. As a result, the smoke entered
unimpeded into the cockpit.
Since the louver panel was missing from the cockpit door,
closing the door would not have blocked totally the entry of the smoke. However, even
with the panel missing, a closed door would have delayed the accumulation of smoke
within the cockpit.
The smoke in the cabin increased rapidly during the emergency descent;
however, the flightcrew did not try to use the cabin smoke elimination’ procedure --
depressurize the airplane after it descended below 10,000 feet, open the aft pressure
bulkhead door, and open slightly the right forward cabin door. The captain testified that
he did not order the procedure to be used because the fire was not out and because he
needed the first officer’s assistance in the cockpit in order to fly the airplane safely.
Had
this procedure been tried, the airflow through the cabin would have been directed forward
and out the right forward cabin door.
Though ideally the procedure would be used when a
fire has been extinguished, according to the airplane manufacturer, the procedure can be
used when a fire is still burning.
According to the manufacturer, the flightcrew must
a
those
”
at
”
’re
”
-
_
_
judge whether survival depends on the elimination of the entrapped smoke regardless of
the effect that a draft of air might have on the fire. However, the fire protection
engineer who investigated the airplane fire damage believed that opening the doors would
have created
very strong that the fire would have been drawn from the
lavatory into the cabin; that the smoke would have been moved forward faster; and that
the use of the procedure would have endangered the passengers and the safety of the
airplane. The evidence concerning what would have happened had the procedure been
used was highly conflicting, and therefore, the Safety Board will not speculate as to what
might have occurred had the procedure been used.
After descending through 3,000 feet, the captain ordered the first officer to
depressurize the airplane in order to prepare it for landing and to insure that the flight
attendants and passengers would be able to open the doors and
exits. The first
officer depressurized the airplane and then, although the procedures do not require this to
be done, turned the air conditioning pack switches off which still would have been
operating because they are powered by the emergency
bus. The evidence showed that
the captain did not order the first officer to shut off the air conditioning packs, nor did he
know that they were shut off. The first officer testified that after the airplane had been
depressurized he decided to turn the packs off because the smoke was getting bad; he
thought he had to do something, and he thought that
packs, they are just feeding
the
The testimony concerning when the air conditioning packs were
shut off is slightly conflicting.
However, the consensus of all the testimony and
statements indicated that they were almost certainly turned off by the time the airplane
reached 2,000 feet.
The ATC transcripts showed that Flight 797 reported
two
thousand five hundred and VFR at and respectively.
The flight landed at
Based on the flightcrew’s testimony and statements and the
times in the ATC transcript, the Safety Board concludes that the air conditioning packs
were turned off at least 4 minutes before the airplane landed in which almost two
complete changes of cabin and cockpit air otherwise would have occurred.
During the descent, the augmentation valves in the air conditioning and
pressurization packs had failed to the closed position; therefore, the smoke in the cabin
and cockpit was not being purged overboard as rapidly as it would have been had these
valves remained operational.
Consequently, the rate at which the smoke was filling the
cabin and cockpit increased which may have led to the first officer’s
that the air
conditioning and pressurization’ packs were supplying air flow to the fire.
Given the
conditions that existed in the cabin during the descent, the first officer’s action was
understandable; however, when the packs were shut off, the best available means of
eliminating smoke from both the cockpit and the cabin was inoperative. Since the ram air
valve never was opened and the pressurization system had been shut off, there was
virtually no fresh air supply to the cockpit and cabin.
With no exchange of air in the
aircraft, smoke and heat continued to accumulate at the ceiling and to build down toward
the floor and the toxicity of the air in the cabin began increasing at an accelerated rate.
Moreover, with no airflow available to reduce the buildup of heat and combustible gases
at the upper levels of the cabin, the onset of factors conducive to flashover or flashfire
was accelerated.
In conclusion, once the decision to descend was made the flightcrew executed
an emergency descent and then, in coordination with the approach controller, flew a
descending traffic pattern to the landing runway using only rudimentary airplane
.
.
‘.
/
‘.
’
‘I.
‘.
‘
”
-
-
_
_
-
“,.
‘“‘“““-“““-
L
instrumentation. The captain, though confronted with a hostile cockpit environment, loss
.
of engine and flight instruments, and an inoperative horizontal stabilizer trim system was
able to maintain his concentration, and with the assistance of his first officer, configure
the airplane for landing, slow it to the desired indicated airspeeds, and despite the
unfamilar longitudinal control forces resulting from the inoperative horizontal stabilizer,
land the airplane safely.
Considering the conditions which confronted the captain during
.
the descent and landing, the Safety Board concludes that the captain exhibited
outstanding airmanship without which the airplane and everyone on board would certainly
,
have perished.
.
.
.
Evacuation of the Airplane
Although fatalities occurred, this accident must be considered survivable
because none of the survivability factors were violated. By definition, a survivable
accident is one in which the forces transmitted to the occupants do not exceed the limits
of human tolerance to abrupt acceleration, either positive or negative, and in which the
structure in the occupant’s immediate environment remains structurally intact to the
extent that an occupiable volume is provided for the occupants throughout the crash
sequence.
In this accident, the fuselage integrity was not breached during the landing
and none
occupants were exposed to decelerative forces beyond the limits of human
tolerances.
According to the passengers and flight attendants, when the airplane landed
the visibility in the cabin was virtually nonexistent at heights higher than 1 foot above the
cabin floor.
In addition, during the descent and landing, the passengers and crew were
exposed to constantly increasing quantities of smoke and toxic gases, and these factors
combined to make the evacuation procedures more difficult to execute and
The flight attendants’ efforts to move the passengers forward of row 12 and
away
the source of smoke and heat (except for two passengers in seats 12D and
12E who had refused to move because their seats
were next to the right forward
emergency exit window) and their selection and briefing of able-bodied male passengers
to open the four
exit windows, apparently were successful since three of these
four exits had been opened and were used by surviving passengers. Additionally, the flight
attendants attempted to brief passengers on how to assume the brace position and other
items relevant to the emergency situation.
However, because of the smoke and toxic
gases in the cabin, they had great difficulty communicating, and in some cases, passengers
did not hear
these instructions. Virtually all the survivors stated that they had covered
their mouths and noses with towels, articles of clothing, or other like items, as instructed
by flight attendants.
Wet towels will filter out smoke particles, acid gases such as
hydrogen chloride and hydrogen fluoride, and hydrogen cyanide. While breathing through
items of clothing will also filter out smoke particles, the clothing would probably be less
effective in filtering out the acid gases and hydrogen cyanide. Neither filter system will
reduce the carbon monoxide concentration. Therefore, although this procedure was not
contained in the company manual, the initiative on the part of the flight attendants to
distribute wet towels and instruct the passengers to breathe through the towels or other
items of clothing may have aided the survival of the passengers.
Aircraft Survival Design Guide, Volume 1:
“Design Criteria and Checklists,” U.S.
Research Technical Laboratory, Technical Report December 1980.
.
behaviorial
’s
In spite of these efforts, several factors limited the success of the evacuation
of the passengers.
The flight attendants at the forward doors were not able to make
themselves heard inside the cabin.
The location of the fatalities in the cabin tends to
confirm that those who succumbed either made no attempt to move toward an exit or
started too late and were overcome as they attempted to move toward an exit. Studies
indicate that in the absence of commands, some passengers will remain seated and await
orders, a phenomenom known as
inaction.
It is also possible that some of
the passengers were incapacitated because of
toxic gases and smoke during
the descent and landing.
The statements and testimony of the survivors who exited the airplane through
the
exit windows indicated that the visibility in this area of the cabin was
probably worse than that in the forward
--The survivors who had moved aft to reach
the
exits found them because they had memorized the number of rows between
their seats and the exits and thereafter counted the rows by feeling the seatbacks as they
moved aft; because they were able to see a dim glow of light when they reached the exit
area; or as in one case, the survivor felt a slight breeze across the back of her legs when
she reached the area of an open exit.
The evidence showed that the cabin environment deteriorated rapidly after the
doors and
exit windows were opened.
Although all of the passengers had been
seated forward of row 13 when the airplane stopped, two of the fatalities were found in
the cabin aisle at rows 14 and 16. It is likely that these passengers had made their way
aft trying to locate the
exits; however, the visibility had deteriorated so badly
that they were not able to locate them. Based on this evidence and the difficulties
experienced by the survivors who were able to locate the
exit windows, the
Safety Board believes that had floor level, or near floor level, emergency lighting denoted
the location of the
emergency exit windows, not only might these two
passengers have been able to find them, the task of the other survivors would have been
made easier.
Many of the survivors stated that they were able to see better either by
bending forward or by crawling. The survivors’ experiences appeared to follow closely the
results of research contained in the AGARD report concerning the stratification of toxic
gas concentrations within a cabin and its effect on survival times. Based on the results of
the
cabin environment research studies, the Safety Board concludes that the cabin
nonsurvivable within 20 to 30 seconds after the flashfire.
The evidence also indicated that there were instances in which the flight
attendants had not complied completely with the Air Canada evacuation procedures. In
the event of failure of the PA system, Air Canada procedures direct the use of a mega-
phone to make required announcements before the airplane door and exit windows are
opened. Even though the flight attendant in charge knew that the airplane PA system was
inoperative, he did not remove the megaphone to make the announcements prescribed in
the company briefing format.
Although he was busy attending to a sick passenger, he
could have required one the other attendants either the passenger or
take the megaphone and make the announcements to augment the individual briefings
given to the passengers by the Nos. 2 and 3 flight attendants. The Safety Board concludes
that had this been done, the emergency briefings probably would have been heard, by
more, if not all, of the passengers, and in any event in greater detail. However, even had
all the passengers heard the briefings, the Safety Board cannot conclude that this would
have altered appreciably the sequence of events which occurred after the doors and exit
windows were opened.
One of the required announcements contained in the briefing
Daniel Johnson, “Behavioral Inaction Under Stress Conditions Similar to the Survivable
Aircraft Accident,” Safety Journal, 1972, First Quarter.
‘.
*
sit
”
”
. .
.
format would have directed the passengers to remain seated while the flight attendants
completed opening the doors and exit windows and until the attendants directed them to
get up and move toward the doors and exits.
The evidence showed that a passenger seated
in row 2 had, in fact, been briefed to restrain the other passengers from moving toward
the forward doors until he received the evacuation command from the flight attendant in
charge. This procedure is designed to prevent congestion at the doors and exits so that
the attendants can operate them without interference.
In a smoke-free environment this
is an excellent procedure; however, on Flight 797 the procedure would have operated to
the detriment of the passengers.
The poor visibility made it impossible for the passengers
to see either the opened doors, the opened exit windows, or the flight attendants. Given
these conditions and given the contents of the briefing. announcement, the Safety Board
believes it highly unlikely that the use of the megaphone during the descent would have
provided a better briefing to the passengers who did not move after the airplane stopped
and the doors and exits were opened.
Since neither the No. 2 or No. 3 flight attendant was able to reach and occupy
her designated emergency landing position,
no direct supervision of the emergency
evacuation through the
exit windows was provided by a flight attendant. Both of
these attendants were in the cabin aisle either briefing or attending passengers When they
were directed to sit down by the first officer.
Despite that, they continued to move along
the aisle checking seatbelts and briefing and comforting passengers. Since the conditions
in the cabin made it impossible for them to look outside and estimate how soon the
airplane would land, it was incumbent upon the flight attendants to seat themselves as
soon as possible after receiving the first officer’s command. The No. 2 and No. 3 flight
attendants were seated in seat 3C and in an aisle seat in rows 7, 8, or 9, respectively,
when the airplane stopped.
Neither flight attendant tried to reach her designated supervisory position at
the
exit windows. order to do so, the No. 2 attendant would have had to have
moved aft against the flow of passengers moving toward the forward doors. The No. 3
flight attendant was closer to the
area. However, had she tried to go aft, it is
doubtful if her presence would have altered the evacuation sequence at these exits. She
probably could not have reached the exits before the passengers seated at rows 10 through
12 reached them, opened them, and left the airplane. The survivors who exited the
airplane through the
exit windows stated that they barely had the strength and
presence of mind to negotiate the exits and that they were 3 to 5 rows closer to the exit
than the flight attendant. Even assuming that the No. 3 flight attendant had reached the
area of the exits, her ability to exercise supervisory functions at the four exit windows
would have been diminished severely, if not totally;
In summary, the evidence showed that two flight attendants had opened the
two forward doors, deployed and inflated the slides, and attempted to call out required
commands and directions to the passengers.
They remained at their posts at those doors
until they were either driven out by the heat or until they believed that no more
passengers were moving toward the doors.
They had briefed passengers on the location
and operation of the
emergency exit windows.
Though not required by company
procedures, they had, until directed by the first officer to passed out wet
towels and directed the recipients to breathe through them and otherwise attended to the
passengers. The Air Canada procedures require their flight attendants to “do all possible
to evacuate everyone, but they are not obliged to risk their own
Given the location
of the flight attendants when the airplane stopped on the runway and given the conditions
within the cabin at that time, the Safety Board can only conclude that any attempt by the
. .
’s
-
flight attendants to move farther aft into the cabin and to remain within the cabin for any
appreciable length of time could not have been made without placing their survival in dire
jeopardy and that it is doubtful that additional lives would have been saved thereby.
2.4
~
The response of the crash-fire-rescue units and the methods used by the
firefighters to extinguish the fire on board Flight 797 were consistent with those used to
fight typical airplane fires, primarily fuel spills and interior fires. The tactics used by
crash-fire-rescue units in response to these types of fires are designed to protect the
passengers leaving the airplane and to assist them in moving from the endangered area.
Thereafter, the attempt can be directed to save the airplane.
The efforts of the
firefighters in this case were complicated by the fact that they did not know how many
passengers and crewmembers were on board and they did not know how much fuel was on
board. However, they did know that smoke had been reported coming from the airplane’s
aft lavatory and that there was smoke or fire in the rear of the airplane.
In evaluating the attempts to attack or contain the fire, the question arises as
to whether an entry at the left forward door with a
would have been possible and
more logical.
From this location, the fire might have been pushed back and several
passengers possibly rescued from the interior.
When the firefighters arrived at the
airplane and began applying foam on the top of the fuselage and on the runway directly
underneath the airplane, flames were not visible.
The fire did not become visible until
after the last survivor had left the airplane.
The on scene commander testified that when
his personnel began the attack, passengers were still exiting through the left forward
door, and he believed that an attempt to enter the airplane at that door at that moment
would have impeded, passenger egress.
He also believed that the deployed escape chute
would have impeded and slowed the entry of his personnel into the airplane. In addition,
the firefighters did not know the seat location of the passengers. Therefore, the first
interior attack was mounted at the left
exit and was designed not only to assist
the passengers in this area, but also to insert firefighters between the fire and the
remainder of the passengers still on board.
The decision to mount the. initial interior
attack through the left
exit was an operational decision made by the on scene
commander, and was based upon his assessment of the situation. Based upon the evidence,
the Safety Board concludes that his decision was reasonable and that the ensuing flashfire
could not have been averted by other tactics.
The evidence also showed that the firefighters involved in the three interior
attacks were not able to don their protective hoods over their self-contained breathing
apparatus. The lack of protective hoods impeded the interior attacks. Testimony of the
airport safety specialist indicated that the requirements for protective equipment
for airport firefighters are ambiguous or nonexistent, and that the provisions of 14 CFR
139 do not specify the need for training in interior firefighting tactics. The deficiencies
of 14 CFR 139 in addressing crash-fire-rescue training and protective equipment have
been addressed in some detail in the National Transportation Safety Board’s recent
Airport Safety Study.
2.5
Fire Prevention/Detection
During this investigation, the Safety Board identified continued presence
of causal factors similar to those identified in previous Safety Board investigations of
Safety Study: Certifications
and
Operations.”
*
-69-
airplane lavatory and cabin fires.
As a result of these earlier investigations, the Safety
Board had made recommendations to the FAA designed either to preclude the recurrence
.
of similar occurrences or to minimize their severity should they recur. Had comprehen-
sive action been taken in response to some of these recommendations, the Safety Board
. .
believes that the severity of this accident would have been lessened. On September 5,
1973,
the Safety Board issued the following Safety Recommendations to the FAA:
.
Require a means for early detection of lavatory fires on all
powered, transport-category aircraft operated under Part 121 of the
Federal Aviation Regulations, such as smoke detectors or operating
procedures for the frequent inspections of lavatories by cabin
attendants. (Safety Recommendation A-73-67)
Require emergency oxygen bottles with full face masks for each cabin
attendant on turbine-powered transport aircraft in order to permit the
attendants to combat lavatory and cabin fires. (Safety Recommendation
A-73-66)
While the FAA has required more frequent inspections by flight attendants, neither smoke
detectors nor full-face smoke masks are required by regulation to be placed on board
transport category aircraft.
Given the conditions inside the lavatory on Flight 797 when
the flight attendants first opened the door, the Safety Board believes that an operable
smoke detector would have alerted the crew to the existence of the smoke before it was
actually discovered.
The flight attendant in charge was able to see the aft wall and that smoke was
emerging from the seams in the lavatory walls when he entered the lavatory. The Safety
Board believes that had an oxygen bottle with a full-face smoke mask been available and
used, it might have encouraged and enabled the first attendant to take immediate and
aggressive actions to locate the source of the smoke and to fight the fire, as set forth in
the company manual. The Safety Board believes
that had either the smoke detector or
the full-face smoke mask, or both, been available and used on Flight 797, the
consequences would have been less severe.
There can be little doubt that, at the very
least, earlier detection of the smoke would have produced a more prompt assessment of
the severity of the conditions in the lavatory, and consequently an earlier decision to
descend and land.
3. CONCLUSIONS
3.1
Findings
1.
The airplane was registered, equipped, and maintained in accordance
. .
with Canadian regulations, and it was operated within the United States
in accordance with applicable Federal Aviation Regulations.
2.
The flightcrew and the cabin crew were qualified and trained in
accordance with Canadian regulations and Air Canada requirements.
Each crewmember had received the off-duty times prescribed
by
Canadian regulations.
3.
A fire propagated through the amenities section of-the aft lavatory and
had burned undetected for almost 15 minutes before the smoke was first
noticed.
.
,
.
.
_
1
3.2
-71-
16.
A flashfire occurred within the cabin within 60 to 90 seconds after the
doors and
window exits were opened. Flames from this’ fire
were not evident until after the survivors had left the airplane. Flames
from the original fire never were evident within the airplane or to
persons on the ground.
17.
This was a survivable accident.
Probable Cause
The National Transportation Safety Board determines that the probable causes
of the accident were a fire of undetermined origin, an underestimate of fire severity, and
misleading fire progress information provided to the captain.
The time taken to evaluate the nature of the fire and to decide to initiate an
emergency descent contributed to the severity of the accident.
4. RECOMMENDATIONS
On July 11, 1973, the Safety Board participated in the investigation of ‘the
Varig Airlines Boeing 707 accident near Paris, France, in which 124 persons died after a
fire erupted in the rear lavatory.
As a result of that accident, the Safety Board, on
September 5, 1973, issued the following Safety Recommendations to the Federal Aviation
Administration (FAA):
Require a means for early detection of lavatory fires on all
turbine-powered, transport-category aircraft operated under Part
121 of the Federal Aviation Regulations, such as smoke detectors
or operating procedures for the frequent inspections of lavatories
by cabin attendants. (Safety Recommendation A-73-67)
Require emergency oxygen bottles with full-face smoke masks for
each cabin attendant on turbine-powered transport aircraft in
order to permit the attendants to combat lavatory and cabin fires.
(Safety Recommendation A-73-68)
Organize a government/industry task force on aircraft fire preven-
tion to review design criteria and formulate specific modifications
for improvements with respect to the fire potential of such
enclosed areas as lavatories in turbine-powered aircraft operating
under the provisions of Part 121 of the
Aviation
Regulations. (Safety Recommendation A-73-70)
Following the investigation of the Pan American World Airways, Inc., Boeing
707 accident that occurred on November 3, 1973, while the flightcrew was attempting to
land at Boston, Massachusetts, after the detection of a fire in the cargo compartment, the
Safety Board issued these additional Safety Recommendations to the FAA:
Provide operators of. the subject aircraft with data to enable
flightcrews to identify smoke sources, and require operators to
procedures in their operating manuals to control and
smoke effectively during the specific flight regimes.
Recommendation A-73-121 issued January 10, 1974)
‘In
’,
-
-
Require that transport category airplanes certificated under Part
4B of the Civil Air Regulations prior to the effective date of
amendment
comply with Part 25.1439 of the Federal Aviation
Regulations. (Safety Recommendation A-74-5, issued February 6,
1974)
Require that a one-time inspection be made of all smoke goggles
provided for the flightcrew of all transport category airplanes to
assure that these goggles conform to the provisions of Part 25.1439
of the Federal Aviation Regulations.
(Safety Recommendation
A-74-6, issued February 6, 1974)
As a result of two other lavatory fires, one aboard a Boeing 747 airplane on
July 17, 1974, and the other aboard a Boeing 727 airplane on August 9, 1974, the Safety
Board recommended that the FAA:
Require that automatic-discharge fire extinguishers be installed in
lavatory waste paper containers on all transport aircraft. (Safety
Recommendation A-74-98, issued December 5, 1974)
In response to Safety Recommendation A-73-67, the FAA issued an Air
Carrier Operations Bulletin (No. 1-76-17,
Flight Lavatory Fires”) instructing Principal
Operations Inspectors to encourage air carriers to prohibit smoking in the lavatories and
l
to institute routine flight attendant inspections of lavatories before takeoff and periodic-
ally during flight. This action was followed by an Airworthiness Directive which required
the installation of “No Smoking” and “No Cigarette Disposal” signs in the lavatories of
transport category airplanes.
Although these actions fell short of the Safety Board’s intention to promote
the installation of smoke detectors such as those using ionization and photo-electric
technology to trigger an alarm signal, the Safety Board, in May 1979, closed Safety
Recommendation A-73-67 and assessed the FAA’s action as acceptable.
While the Safety
Board at that time was sympathetic to the industry’s position that the lack of
demonstrated reliability and the potential for false alarm problems associated with such
smoke detectors would degrade their effectiveness, it now is convinced that the
technology exists to provide an effective and reliable early warning fire detection system
in the lavatories of transport category airplanes. Further, the Safety Board notes that the
FAA report “Feasibility and Tradeoffs of a Transport Fuselage Fire Management System,”
(FAA RD 76-54, dated June 1976) concludes that such systems are feasible with current
technology.
Safety Recommendations A-73-68 and A-74-5 both addressed the need for
standards and requirements for protective breathing equipment to provide flightcrew
members with a supply of oxygen and a mask of eye protection so that they could continue
to perform necessary airplane control functions and cabin duties, as well as firefighting
functions in the event of an in-flight fire.
In response to Safety Recommendation A-73-68, the FAA issued a revision to
the Federal Aviation Regulations effective February
1977, which required the installa-
tion of protective breathing equipment in each isolated separate compartment of the
airplane in which crewmember occupancy is permitted during flight. This revision was not
responsive to the recommendation since it did not provide for portable protective
breathing equipment for use in passenger compartments.’ Also, the FAA issued an NPRM
-
’
-
_’
’s
’s
’s
’s
1
,
-73-
.
in 1975 proposing to amend 14 CFR 25.1439 to include new standards for
oxygen masks
eye coverings.
However, the proposal was later withdrawn with the reasoning that
further testing was needed to establish the standards. The FAA’s response regarding
Safety Recommendation A-73-68 and A-74-5, in August 1981, advised the Safety Board
that an updated Technical Standard Order
would be prepared to prescribe minimum
standards for emergency equipment to provide flightdeck and cabin crewmembers with
eye and respiratory protection from toxic atmospheres during in-flight emergencies.
The
FAA has stated that it intends to issue an Advisory Circular after it adopts the TSO to
recommend
that operators upgrade the protective breathing equipment aboard their
airplanes to meet the new TSO standards. The FAA has stated that the Advisory Circular
would also recommend that operators provide equipment beyond regulatory requirements
for cabin attendants.
The Safety Board assumes that the issuance of
“Protective Breathing Equipment,”
on June 27, 1983, completed the first phase of
intended action. An FAA witness from the Civil Aeromedical Institute testified at the
Safety Board’s public hearing in the Air Canada case that much of the equipment in
current use fails to comply with the newly established minimum standards. He described
serious shortcomings particularly in the effectiveness and fit of smoke goggles. Another
FAA witness from Aviation Standards Office of Airworthiness stated that he was not .
aware of any FAA plans for regulatory action to
that the protective breathing
equipment currently installed on transport category airplanes in accordance with the
provisions of 14 CFR 25.1439 and 14 CFR 121.337 meet the minimum standards prescribed
in
Furthermore, the FAA has not indicated that it intends to require by
regulation the installation of portable breathing equipment which would be available
immediately in passenger compartments for use by cabin attendants in combating cabin
fires. The Safety Board believes that regulatory action is required and that the
contemplated Advisory Circular recommending voluntary action by operators is not
adequate to assure passenger safety.
evaluating the actions regarding Safety Recommendation A-73-70,
the Safety Board acknowledged that the establishment of the Special Aviation Fire and
Explosion Reduction (SAFER) Advisory Committee in May 1978 would be responsive to the
recommendation. The
of this committee, which was composed of government
and industry representatives, went beyond the specific scope of the Safety Recommenda-
tion and considered the broader aspect of the airplane fire problem by addressing the
postcrash scenario. The SAFER committee’s short-term recommendations were directed
primarily toward actions to inhibit the ignition and rapid propagation of a postcrash
fuel-fed fire. The committee determined that there was a need for continued research in
interior cabin materials before new testing procedures and standards could be established
regarding the flammability, smoke, and toxic emission characteristic of cabin materials.
Although the
action to convene the SAFER committee was viewed as
responsive to Safety Recommendation A-73-70, the Safety Board maintained the recom-
mendation in an open status pending further progress toward the mandating of safety
enhancing improvements to airplane cabin interiors. The Safety Board has received no
further response from the FAA regarding this recommendation since March 14, 1979.
However, the Safety Board has followed and has encouraged the continuing research being
conducted at the
Technical Center at Atlantic City, New Jersey. The Safety Board
believes that this research has identified several potential cabin improvements which
could be implemented now.
announced on October 11, 1983,
intention to issue two Notices of Proposed Rulemaking proposing new
performance standards for the use of fire-blocking materials on passenger seats to inhibit
the propagation of cabin fires and new standards for emergency lighting that would
be
more effective for passengers evacuating smoke-filled cabins: One of the parties to the
Air Canada accident investigation has recommended that, in addition relocating the
’s
’s
’s
’s
_
-
.
cabin emergency lights, tactile aisle markers like those on the overhead stowage bins on
many airplanes be installed near to the floor to guide persons to emergency exits in the
smoke-filled environment. The Safety Board agrees with this recommendation. Further,
in addition to the proposed improvements already announced by the Administrator, the
FAA tests conducted at the Technical Center have identified other needed upgrading of
equipment.
These tests have demonstrated vividly that the performance of hand fire
extinguishers with the
extinguishing agent is significantly superior to the
performance of the carbon dioxide, dry chemical, or water-type hand extinguishers and
that safety will be enhanced by replacing the latter types of extinguishers with the
type. The Safety Board strongly encourages the FAA to expedite the rulemaking actions
to make fire-blocking seat materials, improved emergency lighting, tactile emergency
exit indicators, and hand fire extinguishers using advanced technology extinguishing
agents mandatory in the transport airplane fleet as early as practicable.
The FAA acted promptly in response to Safety Recommendation A-73-121 to
assess the adequacy of the smoke removal procedures on the Boeing 707 airplane. As a
result of the
assessment and tests, the relevant section of the airplane’s Flight
Manual was revised to include improved and clearer smoke removal procedures. Both the
recommendation and the
actions were specifically directed to the Boeing
airplane. On that basis, Safety Recommendation A-73-121 was closed and response
was deemed acceptable.
However, the circumstances of the Air Canada accident indicate
that the flightcrew encountered difficulty in controlling smoke in the cockpit of the
McDonnell Douglas DC-9 airplane.
The Safety Board questions the applicability of the
prescribed procedures when a cabin fire continues to generate smoke and toxic gases.
Further, testimony at the public hearing disclosed uncertainties among both flightcrew
and expert witnesses regarding optimal smoke control procedures, such as the best use of
cabin air conditioning systems.
The Safety Board, consequently; believes that smoke
removal procedures in all types of air carrier airplanes should be reassessed.
The FAA did not concur in the Safety Board’s recommendation to require that
automaticdischarge fire extinguishers be installed in lavatory waste receptacles on all
transport airplanes (Safety Recommendation A-74-98). The FAA reasoned that the
combined actions of installing fully sealed waste receptacles to assure fire containment
and extinguishment, as required by Airworthiness Directives for transport category
airplanes, and the prohibition of smoking in airplane lavatories eliminated the need for
mandatory installation of automatic-discharge fire extinguishers.
Although
discharge fire extinguishers have been installed in the lavatory waste receptacles of some
airplanes, including the Air Canada
they have not been required and are not
generally installed.
The Safety Board closed Safety Recommendation 74-98 after
assessing
action as unacceptable.
Moreover, the Safety Board is concerned that the FAA’s actions to assure a
sealed design of the lavatory waste receptacle have not been adequate.
On June 25, 1983,
a flight attendant aboard an Eastern Air Lines McDonnell Douglas DC-9 airplane noticed
smoke coming from the right rear lavatory as the airplane was being taxied to the gate
after landing.
It was determined that the fire had started within the lavatory waste
receptacle and propagated behind the vanity to the lavatory aft wall before it was
extinguished by the airport fire department.
The inspection of the undamaged left rear
lavatory in the airplane revealed that the upper area of the waste chute behind the
disposal door was not sealed to contain a fire, and there was no fire extinguisher in the
receptacle. Further, it was evident that waste could accumulate in the enclosed area of
the vanity adjacent to the waste receptacle.
Following this incident, the Safety Board’s
.
.
’s
’
Issue a Maintenance Alert Bulletin to require Principal Mainten-
ance Inspectors to
that airlines have an acceptable program
for the frequent removal of waste from all areas of the lavatory
with particular attention to those enclosed areas in and around the
waste receptacles, and
which gives sufficient emphasis to areas
susceptible to the accumulation of fluids in the vicinity of wire
harnesses and other electrical components which can cause corro-
sion. (Safety Recommendation A-83-49)
The Safety Board notes that the FAA has, in response to several of the Board’s
Safety Recommendations, issued Notices of Proposed Rule Making
NPRM 83-14,
NPRM 83-15, and
83-14 and 83-15 were issued on October 11, 1983.
The proposed rules contained in NPRM 83-14 establish more stringent flammability
requirements for type certification of transport category airplanes and would require that
previously certified airplanes conform to the more stringent criteria within 3 years from
the date the proposed rules are made effective.
NPRM 83-15 would establish the requirement to provide floor proximity
emergency escape path markings in transport ‘category airplanes.
The proposed rule
requires that the floor proximity emergency escape path provide visual guidance to
passengers when all sources of illumination more than 4 feet above the cabin aisle floor
are obscured by dense smoke.
Previously certified airplanes would have to comply with
the new standard within 2 years from the date the proposed rule becomes effective.
On May 17, 1984, NPRM 84-5, which contains three proposed rules, was issued.
The proposed rules would require the installation of automatic fire extinguishers for each
lavatory disposal receptacle
towels, paper, and waste. The rules would also require
the installation of smoke detector systems in the galleys and lavatories of air transport
category airplanes and increase the number of hand fire extinguishers to be located in
passenger compartments.
Because the chemical agent
1211 has demonstrated
superior performance and effectiveness in combating fires, the proposed rule would
require that at least two
1211 hand fire extinguishers be installed in the airplane
cabins.
All air carriers would have to comply with these provisions within 1 year after the
rules become effective.
Upon its consideration of this accident report, the Safety Board issued the
following additional recommendations to the Federal Aviation Administration:
Require that Air Carrier Principal Operations Inspectors review the
training programs of their respective carriers and if necessary specify
that they be amended to emphasize requirements:
for flightcrews
to take immediate and aggressive action to deter-
mine the source and severity of any reported cabin fire and to
and severity of the fire are not positively and quickly determined
or if immediate extinction is not assured.
for flight attendants to recognize the urgency of informing
crews of the location, source, and severity of any fire or smoke
within the cabin.
.
r
‘a
‘or
for both flightcrews and flight attendants to be knowledgable of
the proper methods of aggressively attacking a cabin fire by
including hands-on-training in the donning of protective breathing
equipment, the use of the fire ax to gain access to the source of
the fire through interior panels which can be penetrated without
risk to essential aircraft components, and the discharge of an
appropriate hand fire extinguisher on an actual fire. (Class II,
Priority Action) (A-84-76)
Require that Airplane Flight Manuals, Air Carrier Flight Operations
Manuals, and Flight Attendant Manuals be amended to include compre-
hensive discussions and illustrations showing the proper use of a fire ax
and the locations in each model of aircraft operated where a fire ax can
be used safely to
access. to fire smoke emission source.
(Class II, Priority Action) (A-84-77)
Require that those interior cabin panels of transport category airplanes,
including panels of the lavatories and galleys, which can be safely
penetrated with a fire ax be identified by an acceptable and standardized
means. (Class II, Priority Action) (A-84-78)
The Safety Board believes that its recommendations. when implemented will reduce
or eliminate possible sources of ignition, provide earlier detection of cabin fires, and
provide improved procedures and equipment for flightcrew and cabin crew personnel to
combat and control cabin fires.
Since these recommendations address every possible fire
source and every possible area where cabin fires most logically could originate, the Safety
Board also believes that its actions will either prevent or reduce the possibility of a
recurrence of a fire similar to that encountered on board Flight 797.
REVISED REPORT ADOPTED
BY THE NATIONAL TRANSPORTATION SAFETY BOARD
JIM BURNETT
Chairman
PATRICIA A. GOLDMAN
Vice Chairman
JOHN LAUBER
Member
January 31,
1986
*This report was revised based on the Safety Board’s reply to a Petition for
Reconsideration of probable cause and findings (See appendix The original report was
adopted on August 8, 1984, by the following members of the National Transportation
Safety Board: Jim Burnett, Chairman; Patricia A. Goldman, Vice Chairman; and G.H.
Patrick Bursley and Vernon L. Grose, Members.
l .
.
’lic
Preceding page blank
-79-
5. APPENDIXES
APPENDIX A
AND PUBLIC
1.
Investigation
The National Transportation Safety Board was notified of the accident about
1930, on June 2, 1983, and immediately dispatched an investigative team to the scene
from its
headquarters.
Investigative groups were formed for operations
and witnesses, air traffic control, meteorology, human factors, structures, powerplants,
systems, flight data recorder, maintenance records, and cockpit voice recorder.
Parties to the investigation were the Federal Aviation Administration, Air
Canada, McDonnell Douglas
United Technologies Corporation, and the
Greater Cincinnati International Airport.
Transport Canada appointed an accredited
representative to assist the Safety Board during the investigation. The accredited
sentative was assisted by advisors from Air Canada, Canadian Air Line Pilots Association,
and Canadian Air Lines Flight Attendants Association.
Transport Canada also made
available its laboratories and laboratory personnel.
2.
Hearing
A
public hearing was held in Fort Mitchell, Kentucky, beginning
August 16, 1983. Parties represented at the hearing were the Federal Aviation
Administration, Air Canada, McDonnell Douglas Corporation, Heath Tecna Corporation,
Canadian Air Line Pilots Association, and the Greater Cincinnati International Airport.
Transport Canada appointed an accredited representative to assist the Safety Board
during the public hearing.
Ihe
-
APPENDIX B
PERSONNEL INFORMATION .
Captain Cameron
Captain Donald S. Cameron, 51, was employed by Air Canada on March 28,
1966.
He holds Canadian Airline Transport Certificate No. 000964 with airplane
single and multiengine land ratings.
The captain is type rated in Grumman G-73,
VC-9, Lockheed L-49, and McDonnell Douglas DC-3, -4, -8, and -9 airplanes.
captain’s last class-l, group-l Canadian Medical Certificate was issued February 17,
1983, and contained the following limitation,
only when required glasses are
available.”
Captain Cameron qualified as captain in the DC-9 during November 1974. He
passed his last proficiency check on February 14, 1983; his last line check on July 16,
1982; and completed his last recurrent training on January 24, 1983. The captain had
flown about 13,000 hours, 4,939 of which were in the PC-9. During the last 90 days,
days, and 24 hours before the accident, he had flown 111 hours, 39 hours, and 9 hours,
respectively. The captain had been off duty about 11 hours 55 minutes before
for this flight. At the time of the accident, he had been on duty about 7 hours 35 minutes,
5 hours 10 minutes of which was flight time.
First Officer Ouimet
First Officer Claude Cuimet, 34, was employed by Air Canada on
November 25, 1973.
He holds Canadian Senior Commercial Pilot Certificate No. ULS
102366 with a Class 1, Group 1 Instrument Rating valid to February 1, 1984; ratings for all
types of Class 7 airplanes of 12,500 pounds or less, and for the
license is not
valid for
in
“airplanes of more than 12,500 pounds engaged in
corn mercial air service and passengers carried.”
His last class-l, group-l Canadian
Medical Certificate was issued April 21, 1983, with no limitations.
First Officer Ouimet qualified as first officer in the DC-9 during February
1979.
He passed his last proficiency check on July 14, 1982; his last line check on May 26,
1983; and completed his last recurrent training on January 12, 1983. The first officer had
flown about 5,650 hours, 2,499 of which were in the During the last 90 days, 30
days, and 24 hours before the accident, he had flown 144 hours, 75 hours, and 9 hours,
respectively.
The first officer had been off duty about 11 hours 55 minutes before
reporting for this flight. At the time of the accident, he had been on duty about 7 hours
35 minutes, 5 hours 10 minutes of which was flight time.
14 CFR 129.15 states:
No person may act as
flight crew-member unless he holds a
current’certificate or license issued or validated by the country in
which that aircraft is registered showing his ability to perform his
duties connected with that aircraft.
Since both the captain and first officer possessed Canadian certificates with DC-9 type
ratings, they were qualified to operate DC-9 type airplanes within the United States.
.
.
.
-
_
APPENDIX C
AIRPLANE INFORMATION
McDonnell Douelas DC-9-32, C-FTLU
The airplane manufacturer’s serial No. 47196 was delivered to Air Canada on
April 7, 1966, and had been operated by the airline continuously since that time. A review
of the airplane’s flight logs and maintenance records showed that all applicable
Airworthiness Directives had been complied with, and that all checks and inspections were
completed within their specified time limits. The records review showed that the airplane
had been maintained in accordance with company procedures and Canadian rules and
regulations and disclosed no discrepancies that could have affected adversely the
performance of the airplane and any of its components.
The airplane was powered by Pratt and Whitney
engines rated at
14,000 pounds of static thrust for takeoff at sea level at 64” F.
The following is pertinent statistical data:
Airplane
Total Airplane Time 36,625 hours
Total Airplane Landings
34,987
Powerplants
Engine
No. 1
No. 2
Serial Number
Total Time
20,942
hours
26,990
hours
Total Cycles
21,459
29,598
’
.
’
CAM
-4
APPENDIX
AIR CANADA COCKPIT VOICE
RECORDER, S/N 1613
REMOVED FROM
DOUGLAS DC-9 WAS INVOLVED
CINCINNATI, OHIO, ON
Cockpit area
microphone voice
or sound source
transmission from
accident aircraft
Voice identified
Voice identified
Flight Attendant
Voice identified female Flight Attendant
Voice identified
as a
male passenger
Indianapolis Center
Varous aircraft
Editorial insertion
times are expressed in central standard
-
INTRA-COCKPIT
AIR-GROUND COMMUNICATIONS
SOURCE
CONTENT
TIME
SOURCE
CONTENT
- -
CAM
((Sound similar to arcing))
1848: 15
CAM
((Sound similar to arcing))
CAM
((Two sounds similar arcing))
CAM-1
How is your sea food, nice?
CAM
((Sounds similar and snapping))
I
CAM-2
It’s good
CAM- 1
steak nice?
CAM- 2
Different, a little bit dry but okay
CAM
((Sounds similar to arcing and snapping))
CAM-2
(What was that?)
I
CAM-1
CAM- 2
It’s right there, I see it
Yeah
CAM- 1
DC bus
CAM-2
Which one is that?
I
’l’R
TIME
SOURCE
CONTENT
CAM- 1
l put it in the book, there
CAM-2
it
CAM- 1
I want to log it, eh
CAM-1
Somebody must have pushed a rag down
the old toilet or something, eh?
1853: 21
CAM- 1
Jammed it, and it overheated
CAM-2
Is it flushing you pushed?
CAM- 1
It’s flushing, yeah
CAM-2
(Motor) *
1853: 30
CAM- 1
Toilet flushing, three
banged
CTR
RDO-2
AIR-GROUND
SOURCE
CONTENT
Air Canada seven ninety seven, contact
Indianapolis on one three three point
zero five
Air Canada seven nine seven, so long
So long
I
SOURCE
CAM-1
CAM-1
I
CAM-2
CAM-1
CAM-2
CAM-2
CAN- 1
CAM-2
CAM-1
CONTENT
Don't see the ground too often, today
eh?
No, a lot of, a lotta cloud eh, the
whole * * * the whole area
Yeah, that feels good
What the does this mean
(Reg a bail)
I don't know
Regional examiner, regional
regional
We may be, I don't know, would be a
three letter code if it was an airport, eh
I don't know, it might be in the, ah,
charts
Regional ---
here's another regional
AIR-GROUND
TIME
SOURCE
CONTENT
RDO-1
CTR
RDO-?
Indianapolis Center, this is Air Canada
seven nine seven maintaining three three
zero direct Louisville on course
Air Canada seven ninety seven Indianapolis
Center roger
***
SOURCE
CAM-1
CAM-2
CAM-1
CAM-l
CAM-1
CAM-1
CAM
CAM-?
CAM-2
CAM-1
CAM-2
AIR-GROUND COMMUNICATIONS
TIME
SOURCE CONTENT
(Well it's)
That (one) is lettered D.G.
Oh I see, oh yeah, yeah *
Alternate, ah, must be our alternate
here
Ah who gives a
Nothing to do with us
((Sound similar to cockpit door))
Yeah thank you sir
I
I
Twenty nine U, W, and X twenty nine, those
are the grid references
Twenty nine, yeah
Twenty nine UWX three --- the left toilet
flushing
Left
-~
TIME
SOURCE
CAM-1
CAM
CAM
1900:oo
CAN
NOTE:
CAM-2
CAN-1
CAM-2
CAM-1
INTRA-COCKPIT
CONTENT
Yeah aft left toilet flush, and they
wouldn't accept a reset
((Sound of first attempt to reset and
sound similar to arcing))
((Sound of second attempt to
sound similar to arcing))
((Sound of third attempt to reset and
sound similar to arcing))
((When questioned later, the captain and
copilot said they did not hear the sound
of arcing noted on the CVR CAM channel.
They said they attempted to reset breakers
just one time each))
Pops as
I
push it
Yeah, right
Yeah
((Sound of cough))
.
Zero two seven set for ya Don
Better ---
have dinner here
((Sound of chime))
AIR-GROUND
TIME
SOURCE
CONTENT
I
’RhCOCKPIT
”
I
AIR-GROUND
TIME 6
SOURCE
TIME
SOURCE
CONTENT
---
Yes
CAM-1
Sergio could I try for mine please
CAM-3
Sure
CAM-1
Thank you very much
CAM-1
Do you want any of that fruit or should
we give it to the girls --- far as I'm
concerned
CAM-2
No
CAM-1
I don't want it
CAM-1
There you go
CAM-2
Thanks
CAM-1
You're in a left turn here to pick up
oh two seven
CAM-2
So okay twenty seven
CAM-1
Louisville to
CAM-1
The next chart yeah that's it
CAR-2
Yeah
We're just over Louisville here
I
I
*
I
TIME
SOURCE
CAM- 2
CAM- 2
1902
CAM-4
CAM- 1
CAM-4
CAM- 2
,
1
CAM- 2
CAM- 1
CAM-4
CAM-S
CAM- ?
CAM-S
CAM- ?
IWTRA-COCKPIT
((Sound of whistling))
Louisville
---
okay
Excuse me, there’s a fire in the washroom
at the back, they’re just oh
went back
to go to put it out
Oh yeah
They’re still, well they’re just gonna
go back now
Want me to go there
Yeah go
l the brakers up
Leave my, leave my, leave
my dinner in
the thing there for a minute
Okay
(Can I buy you a drink cause there’s
going on, drink or a shot)
Ah, I wouldn’t say that
Yeah okay
Still there huh?
Yeah
Cot the, ah, breakers pulled
AIR-GROUND COMJNICATIONS
TIME
SOURCE
CONTENT
’M-COCKPIT
-
’
’s
’d
AIR-GROUND COMJNICATLONS
SOURCE
CONTENT
CAM-I
It’s the motor
Pardon me
CAM- 2
You got all the breakers pulled out?
CAM- 1
The breakers are all pulled, yeah
CAM-4
make ‘em all seat?)
CAM-4
Captain is it okay to move everybody
up as far forward as possible
CAM-2
Okay I eh, you don’t have to do it
now, I can’t go back now,
too
heavy, I think
better go down
CAM-3
I got the passengers seated up front,
you don’t have to worry I think its gonna
be easing up
CAM-2
Okay, it’s starting to clear now
l
TIME
SOURCE
CTR
77L
CTR
CTR
1904 :oo
CTR
. .
,
.
CONTENT
Republic two eighty eight Indianapolis,
Memphis one three three point eight five
three three eight five, goodbye
*
l seven seven lima (Knoxsville)
*
two none zero --
I
Seven seven lima (Knoxsville) roger
Delta sixteen twenty six continue
descent to flight level two four
zero, Indianapolis
Center one two eight five five on
two four zero at twenty eight fifty
five, so long
(Cleared) ah okay
,
.
’IT
’
’s
’ll
TIME
SOURCE
CAM- 1
1904
CAM-1
CAM- 3
CAM- 2
CAM- 1
CAM- ?
CAM-2
CAM- 1
CAM-2
CAM- 1
1904
CAM- 1
CAM- 2
1904
CAM- 1
CAM- 2
CAM- 1
AIR-GROUND
TIME
SOURCE
Okay
Well I want --- hold on then
(Mike) I just can’t go back it too
I will go back
appears better,
okay
Yeah that’s okay
okay, yeah ((simultaneous with
above) )
so
Take the, take the smoke mask
You have control
Take the goggles
I
leave the mask
Okay
((Orignated in back
on
) Okay I want ---
Okay go back whenever you can but
don’t get yourself incapacitated
problem, no problem
.
Okay
((This appears on radio channel
as well))
AIR-GROUND
CONTENT
TIME
SOURCE
CONTENT
286G
Indianapolis good evening Citation two
eight six golf, three one oh
190s : 3s
CAM
((Electrical pulse appears on tape
radio channels))
CTR
Citation two eight six gold Indianapolis
CAM-4
Captain, your first officer wanted me
to tell you that Sergio has put a big
discharge of CO2 in the washroom, it seems
to be subsiding,
right
286G
CTR
286G
Okay we’re proceeding direct Pocket City
Affirmative sir, direct Pocket, direct
Six gold
I
I
1905
B747
Center seven four seven level
four three zero’
CTR
seven four seven Indianapolis roger
RDO- 1
Memphis Center this is Air Canada seven
nine seven
’
‘1...,
‘(’
’
.’
’.
‘,
c
.
. . . .
. .
. .
:
. .
. . .
.
TIME 4
SOURCE
CAM-1
(Coming along okay)
CAM- 3
Getting much better, okay
CAM- 3
I was able to discharge half of the CO2
inside the washroom even though I could
not see the source but its definitely
inside the lavatory.
1
1906:
CAM- 3
Yeah, it’s from the toilet, it’s from
the toilet
CO2 it was almost half a bottle and it
now almost cleared
,
CAM- 1
CAM-3
Okay, thank you
Okay, good luck
CAM
((Sound similar to cockpit door))
CAM-2
Okay, you got it *
CTR
1
AIR-GROUND
TIME
SOURCE
CONTENT
Canada seven ninety seven Indianapolis
Center, go ahead
Yeah, we’ve got an electrical problem
we may be off communication shortly
ah stand by
W
I
I
’ll
IWTRA-COCKPIT
CONTENT
Yeah
Okay
I don’t like what’s happening, I think
we better go down, okay?
Okay
Okay, be back there in a minute
AIR-GROUND
TIME
SOURCE
CONTENT
I
P362
Hello Center, Piedmont three sixty
two, we’re level at flight level
three zero
I
CTR
sixty two Indianapolis
roger
P362
We’ll take direct Holston Mountain
if you can do that
((Recorder goes off))
TIME
‘CAM- 1
CAM- 1
CAM-2
CAM-l
CAM- 2
l
’PENDIX
Ihe
I
,
14 CFR 25 FLAME RESISTANCE
Flame resistance criteria for airplane compartment materials are contained in
14 CFR 25.853. Section 14 CFR
states in part:
Interior ceiling panels, interior wall panels, partitions, galley structure,
large cabinet walls, structural flooring and materials used in the
construction of stowage compartments’ (other than underseat stowage
compartments for stowing small items such as magazines and maps! must
be self-extinguishing when tested vertically in accordance with the
applicable portions of appendix F of this part, or other approved
equivalent methods. The average burn length may not exceed 6 inches
and the average flame time after removal of the flame source may not
exceed 15 seconds. Drippings from the test specimen may not continue
to flame for more than an average of 3 seconds after falling.
Section 14 CFR 25.853(b) states, in part:
Floor covering textiles (including draperies and upholstery) seat cushions,
padding, decorative and nondecorative coated fabrics, leather trays and
galley furnishings, electrical conduit, thermal and acoustical insulation
and insulation covering air
joint and edge covering, cargo
compartment liners, insulation blankets, . . . must be self-extinguishing
when tested vertically in accordance with the applicable portions of
appendix F of this part, or other approved equivalent methods.
average burn length may not exceed 8 inches and the average flame time
after the removal of the flame source may not exceed 15 seconds.
Drippings from the test specimen may not continue to flame for more
than an average of 5 seconds after falling.
The acceptable test procedures,
flame heat, and apparatus required to
demonstrate compliance with 14 CFR 25.653 are contained in appendix F of 14 CFR 25.
With regard to vertical and horizontal flame testing, and the appendix states, in part:
Vertical test. . . . For materials covered by section 25.853(a), the
flame must be applied for 60 seconds and then removed. Flame time,
,
burn length, and flaming time of drippings, if any, must be recorded.
The burn length determined in accordance with paragraph
of this
appendix must be measured to the nearest one-tenth inch.
Horizontal test. . . . The flame must be applied for 15 seconds and
then removed. Minimum of 10 inches of the specimen must be used for
timing purposes, approximately 1
inches must burn before the
burning front reaches the timing zone, and the average burn rate must be
recorded.
Burn length.
Burn length is the distance from the original edge to the
farthest evidence of damage to the test specimen due to flame
impingement,
including areas of partial or complete consumption,
.
charring, or
but not including+ areas sooted, stained,
warped, or discolored, nor areas where material has shrunk-or melted
away from the heat source.
.
’
APPENTMX G
FOR RECONSIDERATION OF
PROBABLE CAUSE
AIR LINE PILOTS ASSOCIATION
MASSACHUSETTS AVENUE. N.W.
WASHINGTON. O.C. 20036
December, 20, 1984
The Honorable James Burnett
Chairman
National Transportation Safety Board
Independence Avenue, S. W.
Washington, D. C.
20594
Dear Burnett:
The
Air Lfne Pilots Association has reviewed the Aircraft Accident Report for
the Air Canada DC-9 accident at
Ohio, on June 2, 1983. We were
quite pleased with the technical effort which was expended by the Safety Board
in their search for the source of the fire which occurred on board the
aircraft.
No stone was left unturned, and while the source could not be
Identified at this time, we believe that eventually the cause will be found.
However,
even though the source could not be located, the results of the
research which was conducted
help to reduce the severity and loss of life
in future accidents of the same type.
Although we concur with the technical sections of the well written report, we
disagree with certain sections of the report which deal with the judgment and
decisions of the flight deck crewmembers.
In this regard,
we are therefore
submitting this Petltion for Reconsideration.
While the Air Line Pilots
has no new evidence at this
time, we believe that the Board did not properly assess the evidence which was
in hand with respect to the time factor In the captain’s decision to descend.
On page 61 of the Board report, the Board states, “Therefore, the Safety Board
realizes that there is a need to evaluate the situation before deciding upon
the emergency action required,
However, in this case, the time to make the
decision appears excessive under the circumstances. . . . The Safety Board
believes that a precautionary emergency descent should have been initiated as
It became evident that the fire had not been visually located and could
be attacked directly with extinguishant.
became known at when the
first officer came forward from his first inspection of the aft cabin area,,
about
3 minutes before the decision to begin an emergency descent.”
.
c
I
”
”
’Okay,
”
”
Okay,
“I
”
“to
-
APPENDIX G
December 20, 1984
James Burnett
wholeheartedly that the emergency descent should have been initiated
as soon as it became evident to the captain that the fire had not been visually
located and could not be combated directly.
However, the evidence presented in
the report indicates that this did not occur at
as stated in the Board
Report.
On page 61 of the report, the Board states, “such reports, (of smoke
in the cabin), often turn out to be smoke from an overheated flushing motor or
waste ignited by a discarded cigarette In a trash receptacle designed to
contain a fire, conditions which are normally Identified and corrected by
flight attendants without further consequences.”
At
the captain, by
his own testimony, and as indicated on page
3 of the Board report,- still
believed the fire to be in the lavatory trash
and he did not decide to
descend at this time because,
expected it (the fire) to be put He had
been led to this decision by a series of positive reports with respect to the
firefighting efforts of the crew.
At
only seconds after the first
officer had suggested that they go down, the flight attendant advised the
captain,
“you don’t have to worry I think it
gonna be easing Seven
seconds later, at
the first officer advised, it’s starting to
clear
At another flight attendant advised, “Captain, your first
officer wanted me to tell you that Serglo has put a big discharge of CO in
‘the
it seems to be subsiding all
Some time after
the flight attendant who gave the first report advised the captain, “Getting
much better, okay.”
At
the flight attendant advised the captain,
“CO
it was almost half a bottle and it now almost cleared.”
At
responded,
thank you.”
Between,
and the
captain received
5 separate communications three different crewmembers
indicating that the fire was being brought under control and that the smoke was
subsiding. It should also be noted that at
39 seconds after the Board
indicates that it was evident to the captain that the fire could not be
attacked directly with extinguishant,
the captain sent the first officer aft to
combat the fire.
He would not have done so had he believed that the fire was
not accessible at that time.
Therefore, the above evidence Indicates that the first time it became evident
to the captain that the fire was not controllable was at
when the
first officer returned from the cabin and advised,
don’t like what’s
happening, I think we better go down, Okay?” Upon being informed for the first
time that the fire was not under control,
the captain made the decision to
begin the descent immediately.
Descent was begun as soon as the first officer
resumed his seat and the aircraft prepared for an emergency descent.
On page 70 of the Board report, under Finding number 8, the Board states,
Thereafter, he misconstrued reports that the fire was abating and he delayed
decision to declare an emergency and descend.” Webster’s dictionary
the word “misconstrue
to mean misinterpret: misunderstand”. The
captain did not misinterpret,
nor
he misunderstand any of the communication
from his crew.
He simply relied upon his crew and believed what he was told.
There
a vast difference between believing an erroneous and
misinterpreting a communication.
The word
carries a connotation
of error on’the
of the captain;
It
that he misinterpreted what
he was told, and thereby
an error. There is simply no evidence to
support such a conclusion. To believe is not to misconstrue.
.
-
“_..
’s
‘s
’s
Due
-
‘.
_
-
’s
’we
’s
-
.
.
APPENDIX
Mr. James Burnett
3
December 1984
The Safety Board
report many more erroneous based upon
misconstrued evidence.
However,
these errors are succinctly described in First
Officer C.
notes on the report. The notes have been
the Safety Board, and we will not reiterate them here. We will
state, however, that we fully support Mr.
conclusions.
In light of the evidence contained fn’the Safety Board’s report, we
respectfully request that in
on page of the report, the
last sentence be changed to read,
to persistent reports from various crew
members that the fire was abating,
he delayed his decision to declare an
emergency and descend.*
The last paragraph of the probable cause states,
“Contributing to the severity
of the accident was the
delayed decision to institute an emergency.
descent.
that statement technically true, it suggests that the
captain’s delayed decision was different from the
that a more prudent
captain would have made under similar circumstances.
We
not believe that to
be true.
The Safety Board has had over to subject the captain’s
decisions and the results of those decisions to minute examination by not only
the Board, but by
staff of many persons.
Considering the not inconsiderable
benefit of hindsight, the time which the Board has had to
analyie pie
circumstances and the number of persons employed in this task, it is
grossly unfair and unrealistic for the Safety
to expect flight crew to
evaluate an incident in
dynamic situation,
and reach the same conclusions in
a few seconds that it took the Safety Board and all of its staff
over to
reach.
might point out that even after having the benefit of all the
and hindsight, during the “Sunshine
meeting of July 20, 1984, not
even the Board Members could agree upon the time when the
decision to
descend should
been made.
We believe that given the constant stream of
positive information which was conveyed to this captain regarding the
severity
of the fire, that prudent captain would made his decision to descend
exactly when this captain did.
Even when
crew performs heroically, this
crew did, the Safety Board cannot seem to resist the temptation to find fault
with the crew,
even though such a finding requires a distortion of the evidence
to support such conclusions. It is
very unfortunate that the news media did
not quote the Board’s comments that the captain exhibited outstanding
airmanship without which the airplane and everyone on board would certainly
have perished, but blared in headlines from coast to coast and in
least two
countries,
delayed decision causes accident.” The crew’s reputation
was therefore destroyed by conclusions reached by the Safety Board which were
not in accordance with the evidence in the Board’s possession.
We therefore respectfully request that the Safety Board change the last
paragraph of the probable cause to indicate that the delayed decision was based
upon evidence which
presented to the captain upon 5 different occasions by
3 different crewmembers, which evidence indicated that the fire was abating
and that the smoke
subsiding,
and that an emergency descent would not be
.
required.
”ao.@
APPENDIX G
Hr. James Burnett
4
December 20,
If we can be of any further assistance in processing this petition, please let
us know.
Sincerely,
Henry A. Duffy, President
HAD/cm
_
-103-
NOTES ON REPORT
AIR CANADA DC-9 C-FTLU
CINCINNATI, JUNE 2, 1983
Submitted by: F/O C. Ouimet
"The N.T.S.B. determines that
probable causes of the accident were a
fire of undetermined origin,
an underestimate of fire severity, and con-
flicting fire progress information provided to the captain. Contributing
to severity of the accident was the flight crew's delayed decision to ins-
titute emergency descent."
So the
criticism
of the crew is contained in a time frame between
and E.D.T., or the first 5 minutes and 32 seconds. Most of the
explanation of that criticism is found on page 57 of the report. Paragraph 2,
half
way says:
he (flight attendant in charge) nor the first officer told the captain
that they had not seen the fire-and that they did not know exactly where it
was or how intensely
was burning."..
The board must have missed what is said by Mr. Benetti on page 13 of Exhi-
bit which is the transcript of the cockpit voice recorder.
.
"I was able to discharge half of the CO2 inside the washroom even though I could
not see
the source but it's definitely the lavatory."
Those are the exact words of Mr. Benetti to captain Cameron.
The report continues with paragraph 3:
"He (first officer) retreated the first time because he did not have smoke
goggles with him...".
In Exhibit first officer statement
"I try to get to the washroom but
could not because of smoke".
In the same N.T.S.B. report on page 3, paragraph 2
it says
The
first officer said that he could not get to the aft lavatory be-
cause the smoke, which had migrated over the last three to four rows of seats,
was too thick".
Also in Exhibit transcript of the voice page 10, I say to-the
captain during my first visit to the cockpit . . . I can't go back now, it's
too heavy".
So I did not retreat and never said I did because .I
not have my smoke goggles,
but because I could not physically walk the washroom in that smoke. I thought
.
.
APPENDIX G
-106-
-4-
l
I
never said not to descend, but I realize it implied just that, and the
captain looked back and saw the cabin clear.
Female flight attendant at seems to be subsiding now".
Flight attendant
in charge at is almost cleared".
That change of s ituation is confirmed by many passengers' statements in
Exhibit 6A.
Arnold Friedman:
"There was a short interval in which the smoke subsided".
Glen W. Davis:
and the smoke cleared for a minute or two".
.
In Exhibit page 39, the Human Factors Group Chairman's factual report:
"Several passengers reported a short interval, estimated by one passenger
to be 5 minutes, during which the smoke subsided.".
I believe that is enough evidence to confirm that at the time we were
ding for a descent, there was a significant change in our situation: a
decrease and complete disappearance of smoke in the cabin. The report
fails to recognize that fact in spite of all the reports
exhibits.
What is called "conflicting fire progress",
"underestimate of fire severity",
reports" were just a simple observation, a noting of the facts
by the crew.
Of course, we did not have the benefit of knowing the future
and we believed what we saw, and so it was for everyone on board.
I remember
when I went back for the second time that some people had resumed their drinking
and talking.
There are a lot of questions that go through your mind in a moment like this:
Was the effective
?
Maybe, the motor- pump blew, produced smoke and died?
Remember the pump was very present in my mind.
Those were reasonable assumptions
at the time, much more than the extraordinary set of circumstances that the
report says happened (page 54, paragraph 5):
"The momentary smoke abatement noted by the first officer and flight attendant
in charge between and (right at the time of our decision to
descend) was probably attributable to the dilution effect of opening the lava-
tory door and discharging the
into the area, reclosing the lavatory door,
and the almost simultaneous failure of the lavatory vent line and the flush
and fill pipe connections and check valve,
all of which increased the venti-
lation rate beneath the toilet shroud and accelerated the flow of smoke and
gases to the area below the lavatory floor and overboard.".
It is the timing and this amazing sequence of events that led us to believe
that the situation had improved, and so it was for everybody on board, and
no one that was not there can pretend otherwise.
Again, all of this is very
compatible with the lines of the report paragraph 4:
.
.
-
_
_
‘
c
APPENDIX
RESPONSE TO PETITION OF
RECONSIDERATION OF CAUSE
National Transportation Safety Board
Washington, D.C. 20594
February 11, 1986
Captain Henry A. Duffy
President
Air Line Pilots Association
Petition for Reconsideration of Probable Cause
Aircraft Accident
Air Canada Flight 797, McDonnell-Douglas
DC-9-32, C-FTLU, Greater Cincinnati International Airport,
Covington, Kentucky, June 2, 1983
Report No.
RESPONSE TO PETITION FOR RECONSIDERATION
.
Based on its review of the Petition for Reconsideration (filed December 20,
the National Transportation Safety Board hereby grants, in part, the Petition for
.
Reconsideration.
In accordance with the Safety Board’s rules (49 CFR Part the Safety Board has
entertained a Petition for Reconsideration of its analysis, findings, and probable cause in
the Aircraft Accident Report:
Air Canada Flight 767, McDonnell-Douglas, DC-9-32,
C-FTLU, Greater Cincinnati International Airport, Covington, Kentucky, June 2, 1983.
On August 8, 1984, the Safety Board issued the report on its investigation of the
accident involving Air Canada Flight 797, a McDonnell-Douglas DC-9-32, of Canadian
Registry, C-FTLU, a regularly scheduled international passenger flight from Dallas,
Texas, to Montreal? Quebec, Canada, with an en route stop at Toronto, Ontario, Canada.
The
left 5 crewmembers and 41 passengers on board. en at
flight level 330 (about 33,000 feet the cabin crew discovered smoke in the left aft
lavatory. After attempting to extinguish the hidden fire and then contacting air traffic
control
and declaring an emergency, the crew made an emergency descent, and
ATC vectored Flight 797 to runway
at the Greater Cincinnati International Airport,
Covington, Kentucky. the pilot stopped the airplane, the airport fire department
began firefighting operations.
Flight attendants and passengers opened the left and right
forward doors, the left forward
exit, and the right forward and aft
exits.
About 60 to 90 seconds after the exits were opened, a flash fire engulfed the
airplane interior. While 18 passengers and 3 flight attendants exited through the forward
doors and slides and the three open
exits to evacuate the airplane, the captain
and first officer exited through their respective cockpit sliding windows.
However,
23 passengers were not able to get out of the plane and died in the fire. The airplane was
destroyed.
As a result of its investigation, the Safety Board determined that
the probable causes of the accident were a fire of undetermined
origin, an underestimate of fire severity, and conflicting fire progress
information provided to the captain.
Contributing to the severity of the accident was the flightcrew’s delayed
decision to institute an emergency descent.
’he
-
APPENDIX
-2-
In its investigation, the Safety Board found sufficient evidence to substantiate the
conclusion that a fire propagated through the amenities section of the aft lavatory and
had burned undetected for almost 15 minutes before the smoke was first noticed.
Although, the Board was not able to identify the origin of the fire, it was able to
eliminate several possibilities and identify its general location. The evidence disclosed
that the fire propagated through
the lower part of the amenities section of the lavatory
vanity. Since the direction of airflow below
the vanity and above the toilet waste tank
permitted venting of smoke, fumes, and hot gases away from the cabin and open area of
the lavatory, the fire went undetected for almost 15 minutes until it began to penetrate
the lavatory liner and rise behind and outboard the liner, subsequently penetrating the
sidewall and ceiling seams of the lavatory liner.
The Petitioner acknowledged that it has no new evidence to present with regard to
the probable cause of the accident.
However,
the Petitioner presents three principal
assertions for the Safety Board’s review: first, the Safety Board did not properly assess
the available evidence regarding the time factor in the captain’s decision to descend and
land the airplane; second, the use of the word “misconstrue” in finding No. 8 carries a
connotation of error on the part of the captain; and third, the contributing statement of
the probable cause suggests that the captain’s decision to institute an emergency descent
was delayed longer than a prudent captain would have waited under similar circumstances.
With regard to the first issue, the Petitioner referenced page 61 of the Board’s
report, which stated, in part, that
Safety Board believes that a precautionary
emergency descent should have been initiated as soon as it became evident that the fire
l
had not been visually located and could not be attacked directly with extinguishant. This
became known at
when the first officer came forward from his first inspection of
the aft cabin about 3 minutes before the decision to begin an emergency descent.”
The Petitioner does not disagree with the Board’s belief,
but contends that the time at
which this conclusion became apparent to the captain did not occur at
and the
Petitioner offers the following points from the Board’s report to support its position:
The Board acknowledged the occurrence of previous in-flight incidents
involving lavatory fires which normally have been identified and
corrected
by flight attendants without further consequences;
Trash receptacles in lavatories are designed to contain a fire;
At
the captain indicated that he believed the fire to be in the
lavatory trash bin and that he did not decide to descend at this time
because he expected the fire to be extinguished;
The captain was misled
by a series of reports from his crew indicating
that the fire was under control and being extinguished. (These were five
reports between
an d
fro m
three
different
crewmembers.)
The captain sent the first officer aft to combat the fire 39 seconds after
the time the Board believes it was evident to the captain that the fire
could not be attacked directly with extinguishant;
.
The captain would not have taken. this action
he believed the fire
was not accessible at that time.
.
fire,”
-
I
APPENDIX H
ACCORDINGLY,
The Petitioner’s Petition for Reconsideration and modification of
probable cause and findings of the airplane accident report on Air
Canada Flight 797,
McDonnell-Douglas DC-9-32, C-FTLU, Greater
Cincinnati International Airport, Covington, Kentucky, June 2, 1983, is
hereby granted in part.
The Board’s original report is revised and a corrected report will be
issued to the public which contains modifications to finding No. 8 and the
probable cause.
The probable cause is revised as follows:
The National Transportation Safety Board determines that the probable
causes of the accident were a fire of undetermined origin, an
underestimate of fire severity, and misleading fire progress information
provided to the captain.
The time taken to evaluate the nature of the fire and to decide to
initiate an emergency descent contributed to the severity of the
accident.
Finding No. 8 is revised to read as follows:
The source of the smoke was never identified either by the flight
attendants or the first officer.
The captain was never told nor did he
inquire as to the precise location and extent of the which had
been reported to him.
Crewmember reports
that the smoke was abating
misled the captain
about the fire severity and he delayed his decision to
declare an emergency and descend.
The Safety Board commends the Petitioner for the preparation of the petition and
.
for its interest in aviation safety.
BURNETT, Chairman,
GOLDMAN, Vice Chairman, and LAUBER, Member,
concurred in the disposition of this Petition for Reconsideration.
*U.S. GOVERNMENT
PRINTING 0FF1CE11986-491-098:20080